From f38ced99843e7d1e8d1c32951c16b39cd8087d74 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Sat, 20 Aug 2022 18:46:03 -0700
Subject: [PATCH 01/48] Script to convert NIST EDF data into BallotLab data.

---
 src/electos/ballotmaker/scripts/README.md | 1 +
 1 file changed, 1 insertion(+)
 create mode 100644 src/electos/ballotmaker/scripts/README.md

diff --git a/src/electos/ballotmaker/scripts/README.md b/src/electos/ballotmaker/scripts/README.md
new file mode 100644
index 0000000..9dcb046
--- /dev/null
+++ b/src/electos/ballotmaker/scripts/README.md
@@ -0,0 +1 @@
+Extract data from NIST EDF models for use by BallotLab.

From 88aa5bc419a0a5ae29f162eea0bb11ae92123b08 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Sat, 20 Aug 2022 20:40:24 -0700
Subject: [PATCH 02/48] First pass at candidate contests.

---
 .../ballotmaker/scripts/ballot-lab-data.py    | 118 ++++++++++++++++++
 1 file changed, 118 insertions(+)
 create mode 100644 src/electos/ballotmaker/scripts/ballot-lab-data.py

diff --git a/src/electos/ballotmaker/scripts/ballot-lab-data.py b/src/electos/ballotmaker/scripts/ballot-lab-data.py
new file mode 100644
index 0000000..49f80e0
--- /dev/null
+++ b/src/electos/ballotmaker/scripts/ballot-lab-data.py
@@ -0,0 +1,118 @@
+from typing import List, Union
+
+from electos.datamodels.nist.models.edf import *
+from electos.datamodels.nist.indexes import ElementIndex
+
+OrderedContent = Union[OrderedContest, OrderedHeader]
+
+
+# --- Utilities
+
+
+def text_content(item):
+    """Return joined lines in internationalized text."""
+    assert isinstance(item, InternationalizedText)
+    text = "\n".join(_.content for _ in item.text)
+    return text
+
+
+def walk_ordered_content(content: List[OrderedContent]):
+    for item in content:
+        if isinstance(item, OrderedContest):
+            yield item
+        elif isinstance(item, OrderedHeader):
+            yield item
+            yield from walk_ordered_content(item.ordered_content)
+        else:
+            raise TypeError(f"Unexpected type: {type(item).__name__}")
+
+
+# --- Ballot Properties
+
+
+def all_ballot_styles(election_report: ElectionReport, index):
+    for ballot_style in index.by_type("BallotStyle"):
+        yield ballot_style
+
+
+def ballot_style_name(ballot_style: BallotStyle):
+    assert len (ballot_style.external_identifier) == 1, \
+        "Not ready to handle multiple BallotStyle external IDs"
+    name = ballot_style.external_identifier[0].value
+    return name
+
+
+def ballot_style_gp_units(ballot_style: BallotStyle, index):
+    for id_ in ballot_style.gp_unit_ids:
+        gp_unit = index.by_id(id_)
+        yield gp_unit
+
+
+def ballot_style_contests(ballot_style: BallotStyle, index):
+    for item in walk_ordered_content(ballot_style.ordered_content):
+        contest = index.by_id(item.contest_id)
+        yield contest
+
+
+def ballot_style_candidate_contests(ballot_style: BallotStyle, index):
+    for contest in ballot_style_contests(ballot_style, index):
+        if not isinstance(contest, CandidateContest):
+            continue
+        candidates = []
+        for selection in contest.contest_selection:
+            assert isinstance(selection, CandidateSelection), \
+                "Unexpected non-candidate selection: {type(selection).__name__}"
+            # Ignore write-ins for tier 1
+            if selection.is_write_in:
+                continue
+            for id_ in selection.candidate_ids:
+                candidate = index.by_id(id_)
+                candidates.append(candidate)
+        yield contest, candidates
+
+
+def candidate_name(candidate: Candidate):
+    name = text_content(candidate.ballot_name)
+    return name
+
+
+# --- Main
+
+import json
+import sys
+from pathlib import Path
+
+
+def report(root, index):
+    ballot_styles = list(all_ballot_styles(root, index))
+    # Only look at the first index
+    ballot_style = ballot_styles[0]
+    name = ballot_style_name(ballot_style)
+    print("name:", name)
+    gp_units = ballot_style_gp_units(ballot_style, index)
+    print("gp units:")
+    for item in gp_units:
+        print(f"- {text_content(item.name)}")
+    print("contests:")
+    contests = ballot_style_candidate_contests(ballot_style, index)
+    for contest, candidates in contests:
+        print(f"- name: {contest.name}")
+        print(f"  type: {contest.vote_variation.value}")
+        print(f"  votes: {contest.votes_allowed}")
+        print(f"  candidates:")
+        for candidate in candidates:
+            print(f"  - {candidate_name(candidate)}")
+
+
+def main():
+    file = Path(sys.argv[1])
+    with file.open() as input:
+        text = input.read()
+        data = json.loads(text)
+    edf = ElectionReport(**data)
+    index = ElementIndex(edf, "ElectionResults")
+    report(edf, index)
+
+
+if __name__ == '__main__':
+    main()

From 41e188cb75879bb996f1c06c29898029fd94f981 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Sun, 21 Aug 2022 12:01:13 -0700
Subject: [PATCH 03/48] Ordered content includes headers; use argparse.

---
 .../ballotmaker/scripts/ballot-lab-data.py    | 39 +++++++++++++++----
 1 file changed, 31 insertions(+), 8 deletions(-)

diff --git a/src/electos/ballotmaker/scripts/ballot-lab-data.py b/src/electos/ballotmaker/scripts/ballot-lab-data.py
index 49f80e0..e725ffe 100644
--- a/src/electos/ballotmaker/scripts/ballot-lab-data.py
+++ b/src/electos/ballotmaker/scripts/ballot-lab-data.py
@@ -16,13 +16,23 @@ def text_content(item):
     return text
 
 
-def walk_ordered_content(content: List[OrderedContent]):
+def walk_ordered_contests(content: List[OrderedContent]):
+    """Walk ordered content yielding contests."""
     for item in content:
         if isinstance(item, OrderedContest):
             yield item
         elif isinstance(item, OrderedHeader):
+            yield from walk_ordered_contests(item.ordered_content)
+        else:
+            raise TypeError(f"Unexpected type: {type(item).__name__}")
+
+
+def walk_ordered_headers(content: List[OrderedContent]):
+    """Walk ordered content yielding headers."""
+    for item in content:
+        if isinstance(item, OrderedHeader):
             yield item
-            yield from walk_ordered_content(item.ordered_content)
+            yield from walk_ordered_headers(item.ordered_content)
         else:
             raise TypeError(f"Unexpected type: {type(item).__name__}")
 
@@ -36,9 +46,12 @@ def all_ballot_styles(election_report: ElectionReport, index):
 
 
 def ballot_style_name(ballot_style: BallotStyle):
-    assert len (ballot_style.external_identifier) == 1, \
-        "Not ready to handle multiple BallotStyle external IDs"
-    name = ballot_style.external_identifier[0].value
+    if ballot_style.external_identifier:
+        assert len(ballot_style.external_identifier) == 1, \
+            "Not ready to handle multiple BallotStyle external IDs"
+        name = ballot_style.external_identifier[0].value
+    else:
+        name = ""
     return name
 
 
@@ -49,7 +62,7 @@ def ballot_style_gp_units(ballot_style: BallotStyle, index):
 
 
 def ballot_style_contests(ballot_style: BallotStyle, index):
-    for item in walk_ordered_content(ballot_style.ordered_content):
+    for item in walk_ordered_contests(ballot_style.ordered_content):
         contest = index.by_id(item.contest_id)
         yield contest
 
@@ -78,8 +91,8 @@ def candidate_name(candidate: Candidate):
 
 # --- Main
 
+import argparse
 import json
-import sys
 from pathlib import Path
 
 
@@ -105,7 +118,17 @@ def report(root, index):
 
 
 def main():
-    file = Path(sys.argv[1])
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        "file", type = Path,
+        help = "Test case data (JSON)"
+    )
+    parser.add_argument(
+        "nth", nargs = "?", type = int, default = 1,
+        help = "Index of the ballot style, starting from 1 (default: 1)"
+    )
+    opts = parser.parse_args()
+    file = opts.file
     with file.open() as input:
         text = input.read()
         data = json.loads(text)

From 25ff3cf685c33f8a1f8edf09a1dd710b70f29355 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Sun, 21 Aug 2022 13:30:33 -0700
Subject: [PATCH 04/48] Flags for keeping write-ins, n-of-m contests.

---
 .../ballotmaker/scripts/ballot-lab-data.py    | 39 ++++++++++++++-----
 1 file changed, 29 insertions(+), 10 deletions(-)

diff --git a/src/electos/ballotmaker/scripts/ballot-lab-data.py b/src/electos/ballotmaker/scripts/ballot-lab-data.py
index e725ffe..6cecf6f 100644
--- a/src/electos/ballotmaker/scripts/ballot-lab-data.py
+++ b/src/electos/ballotmaker/scripts/ballot-lab-data.py
@@ -67,16 +67,21 @@ def ballot_style_contests(ballot_style: BallotStyle, index):
         yield contest
 
 
-def ballot_style_candidate_contests(ballot_style: BallotStyle, index):
+def ballot_style_candidate_contests(
+    ballot_style: BallotStyle, index, keep_write_ins, keep_n_of_m, **opts
+):
     for contest in ballot_style_contests(ballot_style, index):
         if not isinstance(contest, CandidateContest):
             continue
+        # Ignore N-of-M by default
+        if not keep_n_of_m and contest.vote_variation == VoteVariation.N_OF_M:
+            continue
         candidates = []
         for selection in contest.contest_selection:
             assert isinstance(selection, CandidateSelection), \
                 "Unexpected non-candidate selection: {type(selection).__name__}"
-            # Ignore write-ins for tier 1
-            if selection.is_write_in:
+            # Ignore write-ins by default
+            if not keep_write_ins and selection.is_write_in:
                 continue
             for id_ in selection.candidate_ids:
                 candidate = index.by_id(id_)
@@ -96,10 +101,14 @@ def candidate_name(candidate: Candidate):
 from pathlib import Path
 
 
-def report(root, index):
+def report(root, index, nth, **opts):
+    """Generate data needed by BallotLab"""
     ballot_styles = list(all_ballot_styles(root, index))
-    # Only look at the first index
-    ballot_style = ballot_styles[0]
+    if not (1 <= nth <= len(ballot_styles)):
+        print(f"Ballot styles: {nth} is out of range [1-{len(ballot_styles)}]")
+        return
+    nth -= 1
+    ballot_style = ballot_styles[nth]
     name = ballot_style_name(ballot_style)
     print("name:", name)
     gp_units = ballot_style_gp_units(ballot_style, index)
@@ -107,11 +116,11 @@ def report(root, index):
     for item in gp_units:
         print(f"- {text_content(item.name)}")
     print("contests:")
-    contests = ballot_style_candidate_contests(ballot_style, index)
+    contests = ballot_style_candidate_contests(ballot_style, index, **opts)
     for contest, candidates in contests:
         print(f"- name: {contest.name}")
-        print(f"  type: {contest.vote_variation.value}")
-        print(f"  votes: {contest.votes_allowed}")
+        print(f"  vote type: {contest.vote_variation.value}")
+        print(f"  votes allowed: {contest.votes_allowed}")
         print(f"  candidates:")
         for candidate in candidates:
             print(f"  - {candidate_name(candidate)}")
@@ -127,14 +136,24 @@ def main():
         "nth", nargs = "?", type = int, default = 1,
         help = "Index of the ballot style, starting from 1 (default: 1)"
     )
+    parser.add_argument(
+        "--keep-write-ins", action = "store_true",
+        help = "Process write in candidates",
+    )
+    parser.add_argument(
+        "--keep-n-of-m", action = "store_true",
+        help = "Process N-of-M contests",
+    )
     opts = parser.parse_args()
     file = opts.file
+    opts = vars(opts)
+
     with file.open() as input:
         text = input.read()
         data = json.loads(text)
     edf = ElectionReport(**data)
     index = ElementIndex(edf, "ElectionResults")
-    report(edf, index)
+    report(edf, index, **opts)
 
 
 if __name__ == '__main__':

From 584bb410b3c32a5338001112940f312c5c2f6641 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Sun, 21 Aug 2022 15:48:56 -0700
Subject: [PATCH 05/48] Drop 'keep' flags, add '--debug' for exceptions, more
 doc comments.

---
 .../ballotmaker/scripts/ballot-lab-data.py    | 37 ++++++++++---------
 1 file changed, 19 insertions(+), 18 deletions(-)

diff --git a/src/electos/ballotmaker/scripts/ballot-lab-data.py b/src/electos/ballotmaker/scripts/ballot-lab-data.py
index 6cecf6f..5e1d3ac 100644
--- a/src/electos/ballotmaker/scripts/ballot-lab-data.py
+++ b/src/electos/ballotmaker/scripts/ballot-lab-data.py
@@ -10,7 +10,7 @@
 
 
 def text_content(item):
-    """Return joined lines in internationalized text."""
+    """Return joined lines from internationalized text."""
     assert isinstance(item, InternationalizedText)
     text = "\n".join(_.content for _ in item.text)
     return text
@@ -41,6 +41,7 @@ def walk_ordered_headers(content: List[OrderedContent]):
 
 
 def all_ballot_styles(election_report: ElectionReport, index):
+    """Yield all ballot styles."""
     for ballot_style in index.by_type("BallotStyle"):
         yield ballot_style
 
@@ -62,26 +63,25 @@ def ballot_style_gp_units(ballot_style: BallotStyle, index):
 
 
 def ballot_style_contests(ballot_style: BallotStyle, index):
+    """Yield the contests of a ballot style."""
     for item in walk_ordered_contests(ballot_style.ordered_content):
         contest = index.by_id(item.contest_id)
         yield contest
 
 
-def ballot_style_candidate_contests(
-    ballot_style: BallotStyle, index, keep_write_ins, keep_n_of_m, **opts
-):
+def ballot_style_candidate_contests(ballot_style: BallotStyle, index, **opts):
     for contest in ballot_style_contests(ballot_style, index):
         if not isinstance(contest, CandidateContest):
             continue
         # Ignore N-of-M by default
-        if not keep_n_of_m and contest.vote_variation == VoteVariation.N_OF_M:
+        if contest.vote_variation == VoteVariation.N_OF_M:
             continue
         candidates = []
         for selection in contest.contest_selection:
             assert isinstance(selection, CandidateSelection), \
                 "Unexpected non-candidate selection: {type(selection).__name__}"
             # Ignore write-ins by default
-            if not keep_write_ins and selection.is_write_in:
+            if selection.is_write_in:
                 continue
             for id_ in selection.candidate_ids:
                 candidate = index.by_id(id_)
@@ -137,23 +137,24 @@ def main():
         help = "Index of the ballot style, starting from 1 (default: 1)"
     )
     parser.add_argument(
-        "--keep-write-ins", action = "store_true",
-        help = "Process write in candidates",
-    )
-    parser.add_argument(
-        "--keep-n-of-m", action = "store_true",
-        help = "Process N-of-M contests",
+        "--debug", action = "store_true",
+        help = "Enable debugging output and stack traces"
     )
     opts = parser.parse_args()
     file = opts.file
     opts = vars(opts)
 
-    with file.open() as input:
-        text = input.read()
-        data = json.loads(text)
-    edf = ElectionReport(**data)
-    index = ElementIndex(edf, "ElectionResults")
-    report(edf, index, **opts)
+    try:
+        with file.open() as input:
+            text = input.read()
+            data = json.loads(text)
+        edf = ElectionReport(**data)
+        index = ElementIndex(edf, "ElectionResults")
+        report(edf, index, **opts)
+    except Exception as ex:
+        if opts["debug"]:
+            raise ex
+        print("error:", ex)
 
 
 if __name__ == '__main__':

From 43eb184d1de02e1007fe1157bd8bc9deb5fe3d49 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Sun, 21 Aug 2022 16:03:35 -0700
Subject: [PATCH 06/48] Significant refactoring of candidate contests.
 BallotStyles don't drive the contests.

---
 .../ballotmaker/scripts/ballot-lab-data.py    | 109 ++++++++++++------
 1 file changed, 71 insertions(+), 38 deletions(-)

diff --git a/src/electos/ballotmaker/scripts/ballot-lab-data.py b/src/electos/ballotmaker/scripts/ballot-lab-data.py
index 5e1d3ac..3c5eb3a 100644
--- a/src/electos/ballotmaker/scripts/ballot-lab-data.py
+++ b/src/electos/ballotmaker/scripts/ballot-lab-data.py
@@ -3,6 +3,13 @@
 from electos.datamodels.nist.models.edf import *
 from electos.datamodels.nist.indexes import ElementIndex
 
+
+# --- Base Types
+#
+# Schema expresses these as union types not subclasses
+
+Contest = Union[BallotMeasureContest, CandidateContest]
+
 OrderedContent = Union[OrderedContest, OrderedHeader]
 
 
@@ -46,7 +53,8 @@ def all_ballot_styles(election_report: ElectionReport, index):
         yield ballot_style
 
 
-def ballot_style_name(ballot_style: BallotStyle):
+def ballot_style_id(ballot_style: BallotStyle):
+    """Get the text of a ballot style's external identifier if any."""
     if ballot_style.external_identifier:
         assert len(ballot_style.external_identifier) == 1, \
             "Not ready to handle multiple BallotStyle external IDs"
@@ -69,31 +77,64 @@ def ballot_style_contests(ballot_style: BallotStyle, index):
         yield contest
 
 
-def ballot_style_candidate_contests(ballot_style: BallotStyle, index, **opts):
-    for contest in ballot_style_contests(ballot_style, index):
-        if not isinstance(contest, CandidateContest):
-            continue
-        # Ignore N-of-M by default
-        if contest.vote_variation == VoteVariation.N_OF_M:
-            continue
-        candidates = []
-        for selection in contest.contest_selection:
-            assert isinstance(selection, CandidateSelection), \
-                "Unexpected non-candidate selection: {type(selection).__name__}"
-            # Ignore write-ins by default
-            if selection.is_write_in:
-                continue
-            for id_ in selection.candidate_ids:
-                candidate = index.by_id(id_)
-                candidates.append(candidate)
-        yield contest, candidates
-
-
 def candidate_name(candidate: Candidate):
+    """Get the name of a candidate as it appears on a ballot."""
     name = text_content(candidate.ballot_name)
     return name
 
 
+def contest_election_district(contest: Contest, index):
+    """Get the district name of a contest."""
+    district = index.by_id(contest.election_district_id)
+    district = text_content(district.name)
+    return district
+
+
+# Gather & Extract
+#
+# Results are data needed for ballot generation.
+
+def extract_candidate_contest(contest: CandidateContest, index):
+    """Extract candidate contest information needed for ballots."""
+    district = contest_election_district(contest, index)
+    candidates = []
+    write_ins = 0
+    for selection in contest.contest_selection:
+        assert isinstance(selection, CandidateSelection), \
+            f"Unexpected non-candidate selection: {type(selection).__name__}"
+        # Write-ins have no candidate IDs
+        if selection.candidate_ids:
+            for id_ in selection.candidate_ids:
+                candidate = index.by_id(id_)
+                candidates.append(candidate)
+        if selection.is_write_in:
+            write_ins += 1
+    result = {
+        "title": contest.name,
+        "type": "candidate",
+        "vote_type": contest.vote_variation.value,
+        "district": district,
+        "candidates": [candidate_name(_) for _ in candidates],
+        "write_ins": write_ins,
+    }
+    return result
+
+
+def gather_contests(ballot_style: BallotStyle, index):
+    """Extract all contest information needed for ballots."""
+    contests = {
+        kind: [] for kind in ("candidate", "ballot_measure")
+    }
+    for contest in ballot_style_contests(ballot_style, index):
+        if isinstance(contest, CandidateContest):
+            entry = extract_candidate_contest(contest, index)
+            contests["candidate"].append(entry)
+        else:
+            # Ignore other contest types
+            print(f"Skipping contest of type {contest.model__type}")
+    return contests
+
+
 # --- Main
 
 import argparse
@@ -107,23 +148,15 @@ def report(root, index, nth, **opts):
     if not (1 <= nth <= len(ballot_styles)):
         print(f"Ballot styles: {nth} is out of range [1-{len(ballot_styles)}]")
         return
-    nth -= 1
-    ballot_style = ballot_styles[nth]
-    name = ballot_style_name(ballot_style)
-    print("name:", name)
-    gp_units = ballot_style_gp_units(ballot_style, index)
-    print("gp units:")
-    for item in gp_units:
-        print(f"- {text_content(item.name)}")
-    print("contests:")
-    contests = ballot_style_candidate_contests(ballot_style, index, **opts)
-    for contest, candidates in contests:
-        print(f"- name: {contest.name}")
-        print(f"  vote type: {contest.vote_variation.value}")
-        print(f"  votes allowed: {contest.votes_allowed}")
-        print(f"  candidates:")
-        for candidate in candidates:
-            print(f"  - {candidate_name(candidate)}")
+    ballot_style = ballot_styles[nth - 1]
+    data = {}
+    id_ = ballot_style_id(ballot_style)
+    data["ballot_style"] = id_
+    contests = gather_contests(ballot_style, index)
+    if not contests:
+        print(f"No contests found for ballot style: {id_}\n")
+    data["contests"] = contests
+    print(json.dumps(data, indent = 4))
 
 
 def main():

From 24b921f6920d7686ca7bf69c3a696ee439ac782d Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Sun, 21 Aug 2022 16:18:58 -0700
Subject: [PATCH 07/48] Add offices and parties to candidate contests.

---
 .../ballotmaker/scripts/ballot-lab-data.py    | 26 +++++++++++++++++++
 1 file changed, 26 insertions(+)

diff --git a/src/electos/ballotmaker/scripts/ballot-lab-data.py b/src/electos/ballotmaker/scripts/ballot-lab-data.py
index 3c5eb3a..f5d861a 100644
--- a/src/electos/ballotmaker/scripts/ballot-lab-data.py
+++ b/src/electos/ballotmaker/scripts/ballot-lab-data.py
@@ -83,6 +83,28 @@ def candidate_name(candidate: Candidate):
     return name
 
 
+def candidate_contest_offices(contest: CandidateContest, index):
+    """Get any offices associated with a candidate contest."""
+    offices = []
+    if contest.office_ids:
+        for id_ in contest.office_ids:
+            office = index.by_id(id_)
+            name = text_content(office.name)
+            offices.append(name)
+    return offices
+
+
+def candidate_contest_parties(contest: CandidateContest, index):
+    """Get any parties associated with a candidate contest."""
+    parties = []
+    if contest.primary_party_ids:
+        for id_ in contest.primary_party_ids:
+            party = index.by_id(id_)
+            name = text_content(party.name)
+            parties.append(name)
+    return parties
+
+
 def contest_election_district(contest: Contest, index):
     """Get the district name of a contest."""
     district = index.by_id(contest.election_district_id)
@@ -98,6 +120,8 @@ def extract_candidate_contest(contest: CandidateContest, index):
     """Extract candidate contest information needed for ballots."""
     district = contest_election_district(contest, index)
     candidates = []
+    offices = candidate_contest_offices(contest, index)
+    parties = candidate_contest_parties(contest, index)
     write_ins = 0
     for selection in contest.contest_selection:
         assert isinstance(selection, CandidateSelection), \
@@ -115,6 +139,8 @@ def extract_candidate_contest(contest: CandidateContest, index):
         "vote_type": contest.vote_variation.value,
         "district": district,
         "candidates": [candidate_name(_) for _ in candidates],
+        "offices": offices,
+        "parties": parties,
         "write_ins": write_ins,
     }
     return result

From 9f0ab5520aa775b299219ad3ee12cc5289ae01ab Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Sun, 21 Aug 2022 16:45:00 -0700
Subject: [PATCH 08/48] Add ballot measure contests.

---
 .../ballotmaker/scripts/ballot-lab-data.py    | 21 +++++++++++++++++++
 1 file changed, 21 insertions(+)

diff --git a/src/electos/ballotmaker/scripts/ballot-lab-data.py b/src/electos/ballotmaker/scripts/ballot-lab-data.py
index f5d861a..c787ef5 100644
--- a/src/electos/ballotmaker/scripts/ballot-lab-data.py
+++ b/src/electos/ballotmaker/scripts/ballot-lab-data.py
@@ -146,6 +146,24 @@ def extract_candidate_contest(contest: CandidateContest, index):
     return result
 
 
+def extract_ballot_measure_contest(contest: BallotMeasureContest, index):
+    """Extract ballot measure contest information needed for ballots."""
+    choices = []
+    for selection in contest.contest_selection:
+        assert isinstance(selection, BallotMeasureSelection), \
+           f"Unexpected non-ballot measure selection: {type(selection).__name__}"
+        choice = text_content(selection.selection)
+        choices.append(choice)
+    district = contest_election_district(contest, index)
+    result = {
+        "title": contest.name,
+        "type": "ballot measure",
+        "district": district,
+        "choices": choices,
+    }
+    return result
+
+
 def gather_contests(ballot_style: BallotStyle, index):
     """Extract all contest information needed for ballots."""
     contests = {
@@ -155,6 +173,9 @@ def gather_contests(ballot_style: BallotStyle, index):
         if isinstance(contest, CandidateContest):
             entry = extract_candidate_contest(contest, index)
             contests["candidate"].append(entry)
+        elif isinstance(contest, BallotMeasureContest):
+            entry = extract_ballot_measure_contest(contest, index)
+            contests["ballot_measure"].append(entry)
         else:
             # Ignore other contest types
             print(f"Skipping contest of type {contest.model__type}")

From f1fd991d1d22df5e221049a5f636d96502bd0390 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Sun, 21 Aug 2022 16:55:09 -0700
Subject: [PATCH 09/48] Add ballot text; don't return offices or parties for
 candidates.

---
 src/electos/ballotmaker/scripts/ballot-lab-data.py | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/src/electos/ballotmaker/scripts/ballot-lab-data.py b/src/electos/ballotmaker/scripts/ballot-lab-data.py
index c787ef5..22f74c7 100644
--- a/src/electos/ballotmaker/scripts/ballot-lab-data.py
+++ b/src/electos/ballotmaker/scripts/ballot-lab-data.py
@@ -139,8 +139,9 @@ def extract_candidate_contest(contest: CandidateContest, index):
         "vote_type": contest.vote_variation.value,
         "district": district,
         "candidates": [candidate_name(_) for _ in candidates],
-        "offices": offices,
-        "parties": parties,
+        # Leave out offices and parties for now
+        # "offices": offices,
+        # "parties": parties,
         "write_ins": write_ins,
     }
     return result
@@ -155,11 +156,13 @@ def extract_ballot_measure_contest(contest: BallotMeasureContest, index):
         choice = text_content(selection.selection)
         choices.append(choice)
     district = contest_election_district(contest, index)
+    full_text = text_content(contest.full_text)
     result = {
         "title": contest.name,
         "type": "ballot measure",
         "district": district,
         "choices": choices,
+        "text": full_text,
     }
     return result
 

From 309db0eca33f4a25d10fabe83d39f1b4a36b06c4 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Sun, 21 Aug 2022 17:17:24 -0700
Subject: [PATCH 10/48] Add README for ballot-lab-data script.

---
 src/electos/ballotmaker/scripts/README.md | 42 +++++++++++++++++++++++
 1 file changed, 42 insertions(+)

diff --git a/src/electos/ballotmaker/scripts/README.md b/src/electos/ballotmaker/scripts/README.md
index 9dcb046..d30d018 100644
--- a/src/electos/ballotmaker/scripts/README.md
+++ b/src/electos/ballotmaker/scripts/README.md
@@ -1 +1,43 @@
 Extract data from NIST EDF models for use by BallotLab.
+
+## Requirements
+
+- The script requires `nist-datamodels`,  using a version that has element indexes.
+
+## Inputs
+
+Filenames are of the format `{test-case-source}_{ballot-style-id}.json`.
+Note the use of `-` to separate words, and `_` to separate the name parts.
+
+Inputs are EDF JSON files. Get them from:
+
+- https://github.com/TrustTheVote-Project/NIST-1500-100-103-examples/blob/main/test_cases/
+
+The script is `ballot-lab-data.py`. It takes an EDF test case, and the index of
+the ballot style to use (a number from 1 to N, that defaults to 1.).
+
+To run it:
+
+- Install `nist-datamodels`, using a version that has element indexes.
+- Run:
+
+      python ballot-lab-data.py <test-case-file> [<index of ballot style>]
+
+  e.g.
+
+      python ballot-lab-data.py june_test_case.json 1
+
+## Outputs
+
+- Output is JSON files with contests, grouped by contest type.
+- The `VotingVariation` in the EDF is `vote_type` here.
+- Write-ins don't affect the candidate list. They are returned as a count.
+  Presumably they would all be the same and all that's needed is their number.
+  They can be ignored.`
+- The fields were selected to match what is needed for `plurality` candidate
+  contests and a little extra. We can add other `VoteVariation`s and ballot
+  measure contests as needed.
+
+## Complications
+
+- There are no `Header`s or `OrderedHeader`s in the test cases.

From f5090b8e0cd60ab4082c37e3edd4aeb7f2b56ac9 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Sun, 21 Aug 2022 18:10:40 -0700
Subject: [PATCH 11/48] Add party names to candidates.

---
 src/electos/ballotmaker/scripts/ballot-lab-data.py | 12 +++++++++++-
 1 file changed, 11 insertions(+), 1 deletion(-)

diff --git a/src/electos/ballotmaker/scripts/ballot-lab-data.py b/src/electos/ballotmaker/scripts/ballot-lab-data.py
index 22f74c7..28648be 100644
--- a/src/electos/ballotmaker/scripts/ballot-lab-data.py
+++ b/src/electos/ballotmaker/scripts/ballot-lab-data.py
@@ -83,6 +83,13 @@ def candidate_name(candidate: Candidate):
     return name
 
 
+def candidate_party(candidate: Candidate, index):
+    """Get the name of the party of a candidate as it appears on a ballot."""
+    party = index.by_id(candidate.party_id)
+    name = text_content(party.name) if party else ""
+    return name
+
+
 def candidate_contest_offices(contest: CandidateContest, index):
     """Get any offices associated with a candidate contest."""
     offices = []
@@ -138,7 +145,10 @@ def extract_candidate_contest(contest: CandidateContest, index):
         "type": "candidate",
         "vote_type": contest.vote_variation.value,
         "district": district,
-        "candidates": [candidate_name(_) for _ in candidates],
+        "candidates": [
+            { "name": candidate_name(_), "party": candidate_party(_, index) }
+            for _ in candidates
+        ],
         # Leave out offices and parties for now
         # "offices": offices,
         # "parties": parties,

From 55b5b384498739f97b1e44af8b3750bc08a96559 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Tue, 23 Aug 2022 16:28:58 -0700
Subject: [PATCH 12/48] Add votes allowed on candidates races (even if it's the
 default). Fixes BallotLab #82.

---
 src/electos/ballotmaker/scripts/ballot-lab-data.py | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/src/electos/ballotmaker/scripts/ballot-lab-data.py b/src/electos/ballotmaker/scripts/ballot-lab-data.py
index 28648be..3db154c 100644
--- a/src/electos/ballotmaker/scripts/ballot-lab-data.py
+++ b/src/electos/ballotmaker/scripts/ballot-lab-data.py
@@ -144,6 +144,8 @@ def extract_candidate_contest(contest: CandidateContest, index):
         "title": contest.name,
         "type": "candidate",
         "vote_type": contest.vote_variation.value,
+        # Include even when default is 1: don't require caller to track that.
+        "votes_allowed": contest.votes_allowed,
         "district": district,
         "candidates": [
             { "name": candidate_name(_), "party": candidate_party(_, index) }

From a262592b06c874e08ddc0df316fd784443277a0d Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Tue, 23 Aug 2022 16:31:22 -0700
Subject: [PATCH 13/48] Add party abbreviations to candidates. Fixes BallotLab
 #83.

---
 src/electos/ballotmaker/scripts/ballot-lab-data.py | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

diff --git a/src/electos/ballotmaker/scripts/ballot-lab-data.py b/src/electos/ballotmaker/scripts/ballot-lab-data.py
index 3db154c..a3cf618 100644
--- a/src/electos/ballotmaker/scripts/ballot-lab-data.py
+++ b/src/electos/ballotmaker/scripts/ballot-lab-data.py
@@ -84,10 +84,15 @@ def candidate_name(candidate: Candidate):
 
 
 def candidate_party(candidate: Candidate, index):
-    """Get the name of the party of a candidate as it appears on a ballot."""
+    """Get the name and abbreviation of the party of a candidate as it appears on a ballot."""
     party = index.by_id(candidate.party_id)
     name = text_content(party.name) if party else ""
-    return name
+    abbreviation = text_content(party.abbreviation) if party and party.abbreviation else ""
+    result = {
+        "name": name,
+        "abbreviation": abbreviation,
+    }
+    return result
 
 
 def candidate_contest_offices(contest: CandidateContest, index):

From 4a2731604891fc44d7a8e6adbb2819eb9bbc9e34 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Tue, 23 Aug 2022 16:36:06 -0700
Subject: [PATCH 14/48] Include '@id's with contests. Fixes BallotLab #86.

---
 src/electos/ballotmaker/scripts/ballot-lab-data.py | 1 +
 1 file changed, 1 insertion(+)

diff --git a/src/electos/ballotmaker/scripts/ballot-lab-data.py b/src/electos/ballotmaker/scripts/ballot-lab-data.py
index a3cf618..b1ea29d 100644
--- a/src/electos/ballotmaker/scripts/ballot-lab-data.py
+++ b/src/electos/ballotmaker/scripts/ballot-lab-data.py
@@ -146,6 +146,7 @@ def extract_candidate_contest(contest: CandidateContest, index):
         if selection.is_write_in:
             write_ins += 1
     result = {
+        "id": contest.model__id,
         "title": contest.name,
         "type": "candidate",
         "vote_type": contest.vote_variation.value,

From 15efb15ccb9011b7602dbc01c25507c8b0612013 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Tue, 23 Aug 2022 16:48:19 -0700
Subject: [PATCH 15/48] Use 'ContestSelectionId's instead of a count for
 write-ins. Fixes BallotLab #84.

---
 src/electos/ballotmaker/scripts/ballot-lab-data.py | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/src/electos/ballotmaker/scripts/ballot-lab-data.py b/src/electos/ballotmaker/scripts/ballot-lab-data.py
index b1ea29d..1647203 100644
--- a/src/electos/ballotmaker/scripts/ballot-lab-data.py
+++ b/src/electos/ballotmaker/scripts/ballot-lab-data.py
@@ -134,7 +134,7 @@ def extract_candidate_contest(contest: CandidateContest, index):
     candidates = []
     offices = candidate_contest_offices(contest, index)
     parties = candidate_contest_parties(contest, index)
-    write_ins = 0
+    write_ins = []
     for selection in contest.contest_selection:
         assert isinstance(selection, CandidateSelection), \
             f"Unexpected non-candidate selection: {type(selection).__name__}"
@@ -144,7 +144,7 @@ def extract_candidate_contest(contest: CandidateContest, index):
                 candidate = index.by_id(id_)
                 candidates.append(candidate)
         if selection.is_write_in:
-            write_ins += 1
+            write_ins.append(selection.model__id)
     result = {
         "id": contest.model__id,
         "title": contest.name,

From 4355de7e51f87b28b57979e63141269e2c531eaf Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Wed, 24 Aug 2022 13:58:36 -0700
Subject: [PATCH 16/48] Add selection IDs to all candidates, not just
 write-ins. Updates #84.

---
 .../ballotmaker/scripts/ballot-lab-data.py    | 40 +++++++++++--------
 1 file changed, 23 insertions(+), 17 deletions(-)

diff --git a/src/electos/ballotmaker/scripts/ballot-lab-data.py b/src/electos/ballotmaker/scripts/ballot-lab-data.py
index 1647203..cf667fc 100644
--- a/src/electos/ballotmaker/scripts/ballot-lab-data.py
+++ b/src/electos/ballotmaker/scripts/ballot-lab-data.py
@@ -95,6 +95,27 @@ def candidate_party(candidate: Candidate, index):
     return result
 
 
+def candidate_contest_candidates(contest: CandidateContest, index):
+    candidates = []
+    write_ins = []
+    for selection in contest.contest_selection:
+        assert isinstance(selection, CandidateSelection), \
+            f"Unexpected non-candidate selection: {type(selection).__name__}"
+        # Write-ins have no candidate IDs
+        if selection.candidate_ids:
+            for id_ in selection.candidate_ids:
+                candidate = index.by_id(id_)
+                item = {
+                    "id": selection.model__id,
+                    "name": candidate_name(candidate),
+                    "party": candidate_party(candidate, index),
+                }
+                candidates.append(item)
+        if selection.is_write_in:
+            write_ins.append(selection.model__id)
+    return candidates, write_ins
+
+
 def candidate_contest_offices(contest: CandidateContest, index):
     """Get any offices associated with a candidate contest."""
     offices = []
@@ -131,20 +152,9 @@ def contest_election_district(contest: Contest, index):
 def extract_candidate_contest(contest: CandidateContest, index):
     """Extract candidate contest information needed for ballots."""
     district = contest_election_district(contest, index)
-    candidates = []
     offices = candidate_contest_offices(contest, index)
     parties = candidate_contest_parties(contest, index)
-    write_ins = []
-    for selection in contest.contest_selection:
-        assert isinstance(selection, CandidateSelection), \
-            f"Unexpected non-candidate selection: {type(selection).__name__}"
-        # Write-ins have no candidate IDs
-        if selection.candidate_ids:
-            for id_ in selection.candidate_ids:
-                candidate = index.by_id(id_)
-                candidates.append(candidate)
-        if selection.is_write_in:
-            write_ins.append(selection.model__id)
+    candidates, write_ins = candidate_contest_candidates(contest, index)
     result = {
         "id": contest.model__id,
         "title": contest.name,
@@ -153,11 +163,7 @@ def extract_candidate_contest(contest: CandidateContest, index):
         # Include even when default is 1: don't require caller to track that.
         "votes_allowed": contest.votes_allowed,
         "district": district,
-        "candidates": [
-            { "name": candidate_name(_), "party": candidate_party(_, index) }
-            for _ in candidates
-        ],
-        # Leave out offices and parties for now
+        "candidates": candidates,
         # "offices": offices,
         # "parties": parties,
         "write_ins": write_ins,

From 26deeaffb73cf137ef6fbce4bdaa576f6a92289b Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Wed, 24 Aug 2022 15:20:54 -0700
Subject: [PATCH 17/48] Properly set up multi-candidate slates/tickets. Fixes
 BallotLab #85.

This isn't quite right: fails if a slate has candidates from different parties.
Deferring on that case for now.
---
 .../ballotmaker/scripts/ballot-lab-data.py    | 53 ++++++++++++++-----
 1 file changed, 40 insertions(+), 13 deletions(-)

diff --git a/src/electos/ballotmaker/scripts/ballot-lab-data.py b/src/electos/ballotmaker/scripts/ballot-lab-data.py
index cf667fc..cf8ea3f 100644
--- a/src/electos/ballotmaker/scripts/ballot-lab-data.py
+++ b/src/electos/ballotmaker/scripts/ballot-lab-data.py
@@ -1,3 +1,4 @@
+from itertools import groupby
 from typing import List, Union
 
 from electos.datamodels.nist.models.edf import *
@@ -96,23 +97,49 @@ def candidate_party(candidate: Candidate, index):
 
 
 def candidate_contest_candidates(contest: CandidateContest, index):
-    candidates = []
-    write_ins = []
+    """Get candidates for contest, grouped by slate/ticket.
+
+    A slate will collect candidate names together, but have a single ID for
+    the contest selection, and a single party.
+
+    Todo:
+      Handle the case where candidates on a slate don't share a party.
+      There's no clear guarantee of a 1:1 relationship between slates and parties.
+    """
+    candidates_solo = []
+    # Collect individual candidates
     for selection in contest.contest_selection:
         assert isinstance(selection, CandidateSelection), \
             f"Unexpected non-candidate selection: {type(selection).__name__}"
         # Write-ins have no candidate IDs
-        if selection.candidate_ids:
-            for id_ in selection.candidate_ids:
-                candidate = index.by_id(id_)
-                item = {
-                    "id": selection.model__id,
-                    "name": candidate_name(candidate),
-                    "party": candidate_party(candidate, index),
-                }
-                candidates.append(item)
-        if selection.is_write_in:
-            write_ins.append(selection.model__id)
+        if not selection.candidate_ids:
+            continue
+        for id_ in selection.candidate_ids:
+            candidate = index.by_id(id_)
+            candidate = {
+                "id": selection.model__id,
+                "name": candidate_name(candidate),
+                "party": candidate_party(candidate, index),
+            }
+            candidates_solo.append(candidate)
+    # Group candidates by slate.
+    #
+    # Candidates on the same slate will share the 'ContestSelection' ID
+    # Don't try to collect by party.
+    candidates_by_slate = []
+    for _, slate in groupby(candidates_solo, lambda _: _["id"]):
+        slate = list(slate)
+        # Bail out if candidates on a slate don't share a party
+        assert len({_["party"]["name"] for _ in slate}) == 1, \
+            f"Candidates in '{slate[0]['party']['name']}' slate don't all share the same party"
+        candidate = {
+            "id": slate[0]["id"],
+            "name": [candidate["name"] for candidate in slate],
+            "party": slate[0]["party"],
+        }
+        candidates_by_slate.append(candidate)
+    candidates = candidates_by_slate
+    write_ins = [_.model__id for _ in contest.contest_selection if _.is_write_in]
     return candidates, write_ins
 
 

From fb03e3feda903b91555a5a03b1d72c02f49e8e72 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Wed, 24 Aug 2022 16:16:41 -0700
Subject: [PATCH 18/48] Merge write-ins into candidte contests. Mark write-in
 status as boolean (matching the contents of the EDF). No issue attached.

---
 .../ballotmaker/scripts/ballot-lab-data.py    | 38 +++++++++----------
 1 file changed, 18 insertions(+), 20 deletions(-)

diff --git a/src/electos/ballotmaker/scripts/ballot-lab-data.py b/src/electos/ballotmaker/scripts/ballot-lab-data.py
index cf8ea3f..df20475 100644
--- a/src/electos/ballotmaker/scripts/ballot-lab-data.py
+++ b/src/electos/ballotmaker/scripts/ballot-lab-data.py
@@ -106,22 +106,20 @@ def candidate_contest_candidates(contest: CandidateContest, index):
       Handle the case where candidates on a slate don't share a party.
       There's no clear guarantee of a 1:1 relationship between slates and parties.
     """
-    candidates_solo = []
     # Collect individual candidates
+    candidates_solo = []
     for selection in contest.contest_selection:
         assert isinstance(selection, CandidateSelection), \
             f"Unexpected non-candidate selection: {type(selection).__name__}"
-        # Write-ins have no candidate IDs
-        if not selection.candidate_ids:
-            continue
-        for id_ in selection.candidate_ids:
-            candidate = index.by_id(id_)
-            candidate = {
-                "id": selection.model__id,
-                "name": candidate_name(candidate),
-                "party": candidate_party(candidate, index),
-            }
-            candidates_solo.append(candidate)
+        result = {}
+        result["id"] = selection.model__id
+        if selection.candidate_ids:
+            for id_ in selection.candidate_ids:
+                candidate = index.by_id(id_)
+                result["name"] = candidate_name(candidate)
+                result["party"] = candidate_party(candidate, index)
+        result["is_write_in"] = bool(selection.is_write_in)
+        candidates_solo.append(result)
     # Group candidates by slate.
     #
     # Candidates on the same slate will share the 'ContestSelection' ID
@@ -129,18 +127,19 @@ def candidate_contest_candidates(contest: CandidateContest, index):
     candidates_by_slate = []
     for _, slate in groupby(candidates_solo, lambda _: _["id"]):
         slate = list(slate)
-        # Bail out if candidates on a slate don't share a party
-        assert len({_["party"]["name"] for _ in slate}) == 1, \
-            f"Candidates in '{slate[0]['party']['name']}' slate don't all share the same party"
         candidate = {
             "id": slate[0]["id"],
-            "name": [candidate["name"] for candidate in slate],
-            "party": slate[0]["party"],
         }
+        if "party" in slate[0]:
+            assert len({_["party"]["name"] for _ in slate}) == 1, \
+                f"Candidates in '{slate[0]['party']['name']}' slate don't all share the same party"
+            candidate["name"] = [_["name"] for _ in slate]
+            candidate["party"] = slate[0]["party"]
+        candidate["write_in"] = slate[0]["is_write_in"]
         candidates_by_slate.append(candidate)
     candidates = candidates_by_slate
     write_ins = [_.model__id for _ in contest.contest_selection if _.is_write_in]
-    return candidates, write_ins
+    return candidates
 
 
 def candidate_contest_offices(contest: CandidateContest, index):
@@ -181,7 +180,7 @@ def extract_candidate_contest(contest: CandidateContest, index):
     district = contest_election_district(contest, index)
     offices = candidate_contest_offices(contest, index)
     parties = candidate_contest_parties(contest, index)
-    candidates, write_ins = candidate_contest_candidates(contest, index)
+    candidates = candidate_contest_candidates(contest, index)
     result = {
         "id": contest.model__id,
         "title": contest.name,
@@ -193,7 +192,6 @@ def extract_candidate_contest(contest: CandidateContest, index):
         "candidates": candidates,
         # "offices": offices,
         # "parties": parties,
-        "write_ins": write_ins,
     }
     return result
 

From 1547160a9ec71a34a465deb6fca05480faaec1a3 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Sat, 27 Aug 2022 13:23:11 -0700
Subject: [PATCH 19/48] Multi-candidate slates: Fix dropping all candidates
 before the last one. The error was introduced in the process of folding
 write-ins into contests. - Remove convoluted extra step that makes it harder
 to do this correctly. - Candidate names are grouped in arrays. Now do the
 same for parties. - Use 'is_write_in' for the write-in boolean field.

---
 .../ballotmaker/scripts/ballot-lab-data.py    | 53 +++++++++----------
 1 file changed, 24 insertions(+), 29 deletions(-)

diff --git a/src/electos/ballotmaker/scripts/ballot-lab-data.py b/src/electos/ballotmaker/scripts/ballot-lab-data.py
index df20475..64fda85 100644
--- a/src/electos/ballotmaker/scripts/ballot-lab-data.py
+++ b/src/electos/ballotmaker/scripts/ballot-lab-data.py
@@ -99,46 +99,41 @@ def candidate_party(candidate: Candidate, index):
 def candidate_contest_candidates(contest: CandidateContest, index):
     """Get candidates for contest, grouped by slate/ticket.
 
-    A slate will collect candidate names together, but have a single ID for
-    the contest selection, and a single party.
+    A slate has:
+
+    - A single ID for the contest selection
+    - Collects candidate names into an array.
+    - Collects candidate parties into an array.
+
+    Notes:
+        - There's no clear guarantee of a 1:1 relationship between slates and parties.
 
     Todo:
-      Handle the case where candidates on a slate don't share a party.
-      There's no clear guarantee of a 1:1 relationship between slates and parties.
+        - Allow combining shared parties into a single entry.
     """
     # Collect individual candidates
-    candidates_solo = []
+    candidates = []
     for selection in contest.contest_selection:
         assert isinstance(selection, CandidateSelection), \
             f"Unexpected non-candidate selection: {type(selection).__name__}"
-        result = {}
-        result["id"] = selection.model__id
+        names = []
+        parties = []
         if selection.candidate_ids:
             for id_ in selection.candidate_ids:
                 candidate = index.by_id(id_)
-                result["name"] = candidate_name(candidate)
-                result["party"] = candidate_party(candidate, index)
-        result["is_write_in"] = bool(selection.is_write_in)
-        candidates_solo.append(result)
-    # Group candidates by slate.
-    #
-    # Candidates on the same slate will share the 'ContestSelection' ID
-    # Don't try to collect by party.
-    candidates_by_slate = []
-    for _, slate in groupby(candidates_solo, lambda _: _["id"]):
-        slate = list(slate)
-        candidate = {
-            "id": slate[0]["id"],
+                name = candidate_name(candidate)
+                if name:
+                    names.append(name)
+                party = candidate_party(candidate, index)
+                if party:
+                    parties.append(party)
+        result = {
+            "id": selection.model__id,
+            "name": names,
+            "party": parties,
+            "is_write_in": bool(selection.is_write_in)
         }
-        if "party" in slate[0]:
-            assert len({_["party"]["name"] for _ in slate}) == 1, \
-                f"Candidates in '{slate[0]['party']['name']}' slate don't all share the same party"
-            candidate["name"] = [_["name"] for _ in slate]
-            candidate["party"] = slate[0]["party"]
-        candidate["write_in"] = slate[0]["is_write_in"]
-        candidates_by_slate.append(candidate)
-    candidates = candidates_by_slate
-    write_ins = [_.model__id for _ in contest.contest_selection if _.is_write_in]
+        candidates.append(result)
     return candidates
 
 

From 78f933aa6ff66be5efb21c8a6b7e872b9c7c95ba Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Sat, 27 Aug 2022 13:45:43 -0700
Subject: [PATCH 20/48] Multi-candidate slates: Use party IDs to collapse
 identical parties. Handles case where candidates in a slate have different
 parties.

---
 .../ballotmaker/scripts/ballot-lab-data.py      | 17 ++++++++++-------
 1 file changed, 10 insertions(+), 7 deletions(-)

diff --git a/src/electos/ballotmaker/scripts/ballot-lab-data.py b/src/electos/ballotmaker/scripts/ballot-lab-data.py
index 64fda85..5ef0bd4 100644
--- a/src/electos/ballotmaker/scripts/ballot-lab-data.py
+++ b/src/electos/ballotmaker/scripts/ballot-lab-data.py
@@ -86,14 +86,16 @@ def candidate_name(candidate: Candidate):
 
 def candidate_party(candidate: Candidate, index):
     """Get the name and abbreviation of the party of a candidate as it appears on a ballot."""
-    party = index.by_id(candidate.party_id)
+    # Note: party ID is returned to allow de-duplicating parties in callers.
+    id_ = candidate.party_id
+    party = index.by_id(id_)
     name = text_content(party.name) if party else ""
     abbreviation = text_content(party.abbreviation) if party and party.abbreviation else ""
     result = {
         "name": name,
         "abbreviation": abbreviation,
     }
-    return result
+    return result, id_
 
 
 def candidate_contest_candidates(contest: CandidateContest, index):
@@ -104,12 +106,11 @@ def candidate_contest_candidates(contest: CandidateContest, index):
     - A single ID for the contest selection
     - Collects candidate names into an array.
     - Collects candidate parties into an array.
+      - If all candidates in a race share a single party they are combined into
+        one entry in the array.
 
     Notes:
         - There's no clear guarantee of a 1:1 relationship between slates and parties.
-
-    Todo:
-        - Allow combining shared parties into a single entry.
     """
     # Collect individual candidates
     candidates = []
@@ -118,15 +119,17 @@ def candidate_contest_candidates(contest: CandidateContest, index):
             f"Unexpected non-candidate selection: {type(selection).__name__}"
         names = []
         parties = []
+        _party_ids = set()
         if selection.candidate_ids:
             for id_ in selection.candidate_ids:
                 candidate = index.by_id(id_)
                 name = candidate_name(candidate)
                 if name:
                     names.append(name)
-                party = candidate_party(candidate, index)
-                if party:
+                party, _party_id = candidate_party(candidate, index)
+                if party and _party_id not in _party_ids:
                     parties.append(party)
+                    _party_ids.add(_party_id)
         result = {
             "id": selection.model__id,
             "name": names,

From e65864686426cc66e04723c513b336ff8c749f56 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Sat, 27 Aug 2022 13:50:31 -0700
Subject: [PATCH 21/48] Multi-candidate slates: Only collapse party IDs if ALL
 parties are identical.

---
 src/electos/ballotmaker/scripts/ballot-lab-data.py | 10 +++++++---
 1 file changed, 7 insertions(+), 3 deletions(-)

diff --git a/src/electos/ballotmaker/scripts/ballot-lab-data.py b/src/electos/ballotmaker/scripts/ballot-lab-data.py
index 5ef0bd4..29aad1e 100644
--- a/src/electos/ballotmaker/scripts/ballot-lab-data.py
+++ b/src/electos/ballotmaker/scripts/ballot-lab-data.py
@@ -108,6 +108,7 @@ def candidate_contest_candidates(contest: CandidateContest, index):
     - Collects candidate parties into an array.
       - If all candidates in a race share a single party they are combined into
         one entry in the array.
+      - If any candidates differ from the others, parties are listed separately.
 
     Notes:
         - There's no clear guarantee of a 1:1 relationship between slates and parties.
@@ -127,9 +128,12 @@ def candidate_contest_candidates(contest: CandidateContest, index):
                 if name:
                     names.append(name)
                 party, _party_id = candidate_party(candidate, index)
-                if party and _party_id not in _party_ids:
-                    parties.append(party)
-                    _party_ids.add(_party_id)
+                parties.append(party)
+                _party_ids.add(_party_id)
+        # If there's only one party ID, all candidates share the same party.
+        # If there's any divergence track them all individually.
+        if len(_party_ids) == 1:
+            parties = parties[:1]
         result = {
             "id": selection.model__id,
             "name": names,

From d48ddf7f60da822b213695ef7b8ca7c9b6931cd3 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Sat, 27 Aug 2022 13:52:59 -0700
Subject: [PATCH 22/48] Drop null fields of parties; use an empty dictionary if
 a party is missing.

---
 .../ballotmaker/scripts/ballot-lab-data.py     | 18 +++++++++++-------
 1 file changed, 11 insertions(+), 7 deletions(-)

diff --git a/src/electos/ballotmaker/scripts/ballot-lab-data.py b/src/electos/ballotmaker/scripts/ballot-lab-data.py
index 29aad1e..a7306e1 100644
--- a/src/electos/ballotmaker/scripts/ballot-lab-data.py
+++ b/src/electos/ballotmaker/scripts/ballot-lab-data.py
@@ -85,16 +85,20 @@ def candidate_name(candidate: Candidate):
 
 
 def candidate_party(candidate: Candidate, index):
-    """Get the name and abbreviation of the party of a candidate as it appears on a ballot."""
+    """Get the name and abbreviation of the party of a candidate as it appears on a ballot.
+
+       Drop either field from result if it isn't present.
+    """
     # Note: party ID is returned to allow de-duplicating parties in callers.
     id_ = candidate.party_id
     party = index.by_id(id_)
-    name = text_content(party.name) if party else ""
-    abbreviation = text_content(party.abbreviation) if party and party.abbreviation else ""
-    result = {
-        "name": name,
-        "abbreviation": abbreviation,
-    }
+    name = text_content(party.name) if party else None
+    abbreviation = text_content(party.abbreviation) if party and party.abbreviation else None
+    result = {}
+    if name:
+        result["name"] = name
+    if abbreviation:
+        result["abbreviation"] = abbreviation
     return result, id_
 
 

From 4d1bfb4188954172848c7ced94e1e9bdc97d7e84 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Sat, 27 Aug 2022 15:17:08 -0700
Subject: [PATCH 23/48] Add IDs to ballot measures contests.

---
 src/electos/ballotmaker/scripts/ballot-lab-data.py | 1 +
 1 file changed, 1 insertion(+)

diff --git a/src/electos/ballotmaker/scripts/ballot-lab-data.py b/src/electos/ballotmaker/scripts/ballot-lab-data.py
index a7306e1..0ddb5d5 100644
--- a/src/electos/ballotmaker/scripts/ballot-lab-data.py
+++ b/src/electos/ballotmaker/scripts/ballot-lab-data.py
@@ -213,6 +213,7 @@ def extract_ballot_measure_contest(contest: BallotMeasureContest, index):
     district = contest_election_district(contest, index)
     full_text = text_content(contest.full_text)
     result = {
+        "id": contest.model__id,
         "title": contest.name,
         "type": "ballot measure",
         "district": district,

From 8fd93373d252b55ece4fcaf495144e6b72945889 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Mon, 29 Aug 2022 16:48:37 -0700
Subject: [PATCH 24/48] Add IDs to ballot measure contest selections. Fixes
 BallotLab #86.

---
 src/electos/ballotmaker/scripts/ballot-lab-data.py | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/src/electos/ballotmaker/scripts/ballot-lab-data.py b/src/electos/ballotmaker/scripts/ballot-lab-data.py
index 0ddb5d5..a1e1718 100644
--- a/src/electos/ballotmaker/scripts/ballot-lab-data.py
+++ b/src/electos/ballotmaker/scripts/ballot-lab-data.py
@@ -209,7 +209,10 @@ def extract_ballot_measure_contest(contest: BallotMeasureContest, index):
         assert isinstance(selection, BallotMeasureSelection), \
            f"Unexpected non-ballot measure selection: {type(selection).__name__}"
         choice = text_content(selection.selection)
-        choices.append(choice)
+        choices.append({
+            "id": selection.model__id,
+            "choice": choice,
+        })
     district = contest_election_district(contest, index)
     full_text = text_content(contest.full_text)
     result = {

From 6a88c509d40aa6c21d52531b3dd9ac93c6f29836 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Tue, 30 Aug 2022 16:45:08 -0700
Subject: [PATCH 25/48] Update README for 'ballot-lab-data' script.

---
 src/electos/ballotmaker/scripts/README.md | 49 ++++++++++++-----------
 1 file changed, 25 insertions(+), 24 deletions(-)

diff --git a/src/electos/ballotmaker/scripts/README.md b/src/electos/ballotmaker/scripts/README.md
index d30d018..f840ce5 100644
--- a/src/electos/ballotmaker/scripts/README.md
+++ b/src/electos/ballotmaker/scripts/README.md
@@ -1,43 +1,44 @@
-Extract data from NIST EDF models for use by BallotLab.
+Data to use for BallotLab inputs. The data is extracted from EDF test cases
+and constrained to the data model Ballot Lab is using.
 
-## Requirements
+Output file naming format is `{test-case-source}_{ballot-style-id}.json`.
+Note the use of `-` to separate words, and `_` to separate the name parts.
 
-- The script requires `nist-datamodels`,  using a version that has element indexes.
+All the current examples are taken from these EDF files:
 
-## Inputs
+- https://github.com/TrustTheVote-Project/NIST-1500-100-103-examples/blob/main/test_cases/june_test_case.json
 
-Filenames are of the format `{test-case-source}_{ballot-style-id}.json`.
-Note the use of `-` to separate words, and `_` to separate the name parts.
+To run it:
 
-Inputs are EDF JSON files. Get them from:
+- Install the BallotLab fork and change to the 'edf-data-to-ballot' branch.
 
-- https://github.com/TrustTheVote-Project/NIST-1500-100-103-examples/blob/main/test_cases/
+      git clone https://github.com/ion-oset/BallotLab -b edf-data-to-ballot
 
-The script is `ballot-lab-data.py`. It takes an EDF test case, and the index of
-the ballot style to use (a number from 1 to N, that defaults to 1.).
+- Install the project dependencies:
 
-To run it:
+      poetry install
 
-- Install `nist-datamodels`, using a version that has element indexes.
 - Run:
 
-      python ballot-lab-data.py <test-case-file> [<index of ballot style>]
+      python scripts/ballot-lab-data.py <test-case-file> <index of ballot style>
 
   e.g.
 
-      python ballot-lab-data.py june_test_case.json 1
+      python scripts/ballot-lab-data.py june_test_case.json 1
 
-## Outputs
+Structure of output:
 
-- Output is JSON files with contests, grouped by contest type.
-- The `VotingVariation` in the EDF is `vote_type` here.
-- Write-ins don't affect the candidate list. They are returned as a count.
-  Presumably they would all be the same and all that's needed is their number.
-  They can be ignored.`
+- Output is a series of contests, grouped by contest type (candidate, ballot
+  measure)
+- Within a contest type order of records is preserved.
+- The `VotingVariation` in the EDF is `vote_type` here. It can be filtered.
+  - `vote_type` of `plurality` is the simplest kind of ballot.
+  - Ignore `n-of-m` and `ranked-choice` until later.
+- Write-ins are integrated into the candidate list.
 - The fields were selected to match what is needed for `plurality` candidate
-  contests and a little extra. We can add other `VoteVariation`s and ballot
-  measure contests as needed.
+  contests and a little extra.
+  - To add fields or modify them we should modify `extract_{contest type}_contest`.
 
-## Complications
+Notes:
 
-- There are no `Header`s or `OrderedHeader`s in the test cases.
+- There are no `Header`s or `OrderedHeader`s in `june_test_case.json`.

From 827b2c9efdf21e6394d66b3873c2ae952c67a64a Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Fri, 9 Sep 2022 15:46:53 -0700
Subject: [PATCH 26/48] Move 'scripts' to 'data' and rename script.

---
 src/electos/ballotmaker/{scripts => data}/README.md          | 0
 src/electos/ballotmaker/{scripts => data}/ballot-lab-data.py | 0
 2 files changed, 0 insertions(+), 0 deletions(-)
 rename src/electos/ballotmaker/{scripts => data}/README.md (100%)
 rename src/electos/ballotmaker/{scripts => data}/ballot-lab-data.py (100%)

diff --git a/src/electos/ballotmaker/scripts/README.md b/src/electos/ballotmaker/data/README.md
similarity index 100%
rename from src/electos/ballotmaker/scripts/README.md
rename to src/electos/ballotmaker/data/README.md
diff --git a/src/electos/ballotmaker/scripts/ballot-lab-data.py b/src/electos/ballotmaker/data/ballot-lab-data.py
similarity index 100%
rename from src/electos/ballotmaker/scripts/ballot-lab-data.py
rename to src/electos/ballotmaker/data/ballot-lab-data.py

From 8d6d4318f2ee0d5ad01c2b64e569d97dc31190e0 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Fri, 9 Sep 2022 15:55:35 -0700
Subject: [PATCH 27/48] Separate EDF to data into a script and a module.

---
 .../ballotmaker/data/edf-to-ballot-data.py    | 64 +++++++++++++++++++
 .../data/{ballot-lab-data.py => extractor.py} | 62 +-----------------
 2 files changed, 65 insertions(+), 61 deletions(-)
 create mode 100644 src/electos/ballotmaker/data/edf-to-ballot-data.py
 rename src/electos/ballotmaker/data/{ballot-lab-data.py => extractor.py} (83%)

diff --git a/src/electos/ballotmaker/data/edf-to-ballot-data.py b/src/electos/ballotmaker/data/edf-to-ballot-data.py
new file mode 100644
index 0000000..54bc04e
--- /dev/null
+++ b/src/electos/ballotmaker/data/edf-to-ballot-data.py
@@ -0,0 +1,64 @@
+import argparse
+import json
+from pathlib import Path
+
+from electos.datamodels.nist.indexes import ElementIndex
+from electos.datamodels.nist.models.edf import ElectionReport
+
+from electos.ballotmaker.data.extractor import (
+    all_ballot_styles,
+    ballot_style_id,
+    gather_contests,
+)
+
+
+def report(root, index, nth, **opts):
+    """Generate data needed by BallotLab."""
+    ballot_styles = list(all_ballot_styles(root, index))
+    if not (1 <= nth <= len(ballot_styles)):
+        print(f"Ballot styles: {nth} is out of range [1-{len(ballot_styles)}]")
+        return
+    ballot_style = ballot_styles[nth - 1]
+    data = {}
+    id_ = ballot_style_id(ballot_style)
+    data["ballot_style"] = id_
+    contests = gather_contests(ballot_style, index)
+    if not contests:
+        print(f"No contests found for ballot style: {id_}\n")
+    data["contests"] = contests
+    print(json.dumps(data, indent = 4))
+
+
+def main():
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        "file", type = Path,
+        help = "Test case data (JSON)"
+    )
+    parser.add_argument(
+        "nth", nargs = "?", type = int, default = 1,
+        help = "Index of the ballot style, starting from 1 (default: 1)"
+    )
+    parser.add_argument(
+        "--debug", action = "store_true",
+        help = "Enable debugging output and stack traces"
+    )
+    opts = parser.parse_args()
+    file = opts.file
+    opts = vars(opts)
+
+    try:
+        with file.open() as input:
+            text = input.read()
+            data = json.loads(text)
+        edf = ElectionReport(**data)
+        index = ElementIndex(edf, "ElectionResults")
+        report(edf, index, **opts)
+    except Exception as ex:
+        if opts["debug"]:
+            raise ex
+        print("error:", ex)
+
+
+if __name__ == '__main__':
+    main()
diff --git a/src/electos/ballotmaker/data/ballot-lab-data.py b/src/electos/ballotmaker/data/extractor.py
similarity index 83%
rename from src/electos/ballotmaker/data/ballot-lab-data.py
rename to src/electos/ballotmaker/data/extractor.py
index a1e1718..1c94274 100644
--- a/src/electos/ballotmaker/data/ballot-lab-data.py
+++ b/src/electos/ballotmaker/data/extractor.py
@@ -1,8 +1,7 @@
-from itertools import groupby
 from typing import List, Union
 
-from electos.datamodels.nist.models.edf import *
 from electos.datamodels.nist.indexes import ElementIndex
+from electos.datamodels.nist.models.edf import *
 
 
 # --- Base Types
@@ -242,62 +241,3 @@ def gather_contests(ballot_style: BallotStyle, index):
             # Ignore other contest types
             print(f"Skipping contest of type {contest.model__type}")
     return contests
-
-
-# --- Main
-
-import argparse
-import json
-from pathlib import Path
-
-
-def report(root, index, nth, **opts):
-    """Generate data needed by BallotLab"""
-    ballot_styles = list(all_ballot_styles(root, index))
-    if not (1 <= nth <= len(ballot_styles)):
-        print(f"Ballot styles: {nth} is out of range [1-{len(ballot_styles)}]")
-        return
-    ballot_style = ballot_styles[nth - 1]
-    data = {}
-    id_ = ballot_style_id(ballot_style)
-    data["ballot_style"] = id_
-    contests = gather_contests(ballot_style, index)
-    if not contests:
-        print(f"No contests found for ballot style: {id_}\n")
-    data["contests"] = contests
-    print(json.dumps(data, indent = 4))
-
-
-def main():
-    parser = argparse.ArgumentParser()
-    parser.add_argument(
-        "file", type = Path,
-        help = "Test case data (JSON)"
-    )
-    parser.add_argument(
-        "nth", nargs = "?", type = int, default = 1,
-        help = "Index of the ballot style, starting from 1 (default: 1)"
-    )
-    parser.add_argument(
-        "--debug", action = "store_true",
-        help = "Enable debugging output and stack traces"
-    )
-    opts = parser.parse_args()
-    file = opts.file
-    opts = vars(opts)
-
-    try:
-        with file.open() as input:
-            text = input.read()
-            data = json.loads(text)
-        edf = ElectionReport(**data)
-        index = ElementIndex(edf, "ElectionResults")
-        report(edf, index, **opts)
-    except Exception as ex:
-        if opts["debug"]:
-            raise ex
-        print("error:", ex)
-
-
-if __name__ == '__main__':
-    main()

From 2bec0c11c89ee68a76d207d6864ae55536d0ca99 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Fri, 9 Sep 2022 15:58:35 -0700
Subject: [PATCH 28/48] Use explicit EDF model imports ballot data extractor.

---
 src/electos/ballotmaker/data/extractor.py | 13 ++++++++++++-
 1 file changed, 12 insertions(+), 1 deletion(-)

diff --git a/src/electos/ballotmaker/data/extractor.py b/src/electos/ballotmaker/data/extractor.py
index 1c94274..7649630 100644
--- a/src/electos/ballotmaker/data/extractor.py
+++ b/src/electos/ballotmaker/data/extractor.py
@@ -1,7 +1,18 @@
 from typing import List, Union
 
 from electos.datamodels.nist.indexes import ElementIndex
-from electos.datamodels.nist.models.edf import *
+from electos.datamodels.nist.models.edf import (
+    BallotMeasureContest,
+    BallotMeasureSelection,
+    BallotStyle,
+    Candidate,
+    CandidateContest,
+    CandidateSelection,
+    ElectionReport,
+    InternationalizedText,
+    OrderedContest,
+    OrderedHeader,
+)
 
 
 # --- Base Types

From 05fac8ffe6161dd159cf9f2b4e3a882cd00a216e Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Fri, 9 Sep 2022 16:04:48 -0700
Subject: [PATCH 29/48] Simplify naming of ballot data extractor functions.

---
 .../ballotmaker/data/edf-to-ballot-data.py    | 12 ++++-----
 src/electos/ballotmaker/data/extractor.py     | 26 +++++++++----------
 2 files changed, 19 insertions(+), 19 deletions(-)

diff --git a/src/electos/ballotmaker/data/edf-to-ballot-data.py b/src/electos/ballotmaker/data/edf-to-ballot-data.py
index 54bc04e..a7d0634 100644
--- a/src/electos/ballotmaker/data/edf-to-ballot-data.py
+++ b/src/electos/ballotmaker/data/edf-to-ballot-data.py
@@ -6,23 +6,23 @@
 from electos.datamodels.nist.models.edf import ElectionReport
 
 from electos.ballotmaker.data.extractor import (
-    all_ballot_styles,
-    ballot_style_id,
-    gather_contests,
+    ballot_style_external_id,
+    extract_ballot_styles,
+    extract_contests,
 )
 
 
 def report(root, index, nth, **opts):
     """Generate data needed by BallotLab."""
-    ballot_styles = list(all_ballot_styles(root, index))
+    ballot_styles = list(extract_ballot_styles(root, index))
     if not (1 <= nth <= len(ballot_styles)):
         print(f"Ballot styles: {nth} is out of range [1-{len(ballot_styles)}]")
         return
     ballot_style = ballot_styles[nth - 1]
     data = {}
-    id_ = ballot_style_id(ballot_style)
+    id_ = ballot_style_external_id(ballot_style)
     data["ballot_style"] = id_
-    contests = gather_contests(ballot_style, index)
+    contests = extract_contests(ballot_style, index)
     if not contests:
         print(f"No contests found for ballot style: {id_}\n")
     data["contests"] = contests
diff --git a/src/electos/ballotmaker/data/extractor.py b/src/electos/ballotmaker/data/extractor.py
index 7649630..c0bf0b9 100644
--- a/src/electos/ballotmaker/data/extractor.py
+++ b/src/electos/ballotmaker/data/extractor.py
@@ -58,13 +58,7 @@ def walk_ordered_headers(content: List[OrderedContent]):
 # --- Ballot Properties
 
 
-def all_ballot_styles(election_report: ElectionReport, index):
-    """Yield all ballot styles."""
-    for ballot_style in index.by_type("BallotStyle"):
-        yield ballot_style
-
-
-def ballot_style_id(ballot_style: BallotStyle):
+def ballot_style_external_id(ballot_style: BallotStyle):
     """Get the text of a ballot style's external identifier if any."""
     if ballot_style.external_identifier:
         assert len(ballot_style.external_identifier) == 1, \
@@ -187,12 +181,12 @@ def contest_election_district(contest: Contest, index):
     return district
 
 
-# Gather & Extract
+# --- Extraction
 #
-# Results are data needed for ballot generation.
+# Gather and select data needed for ballot generation.
 
 def extract_candidate_contest(contest: CandidateContest, index):
-    """Extract candidate contest information needed for ballots."""
+    """Extract candidate contest subset needed for a ballot."""
     district = contest_election_district(contest, index)
     offices = candidate_contest_offices(contest, index)
     parties = candidate_contest_parties(contest, index)
@@ -213,7 +207,7 @@ def extract_candidate_contest(contest: CandidateContest, index):
 
 
 def extract_ballot_measure_contest(contest: BallotMeasureContest, index):
-    """Extract ballot measure contest information needed for ballots."""
+    """Extract ballot measure contest subset needed for a ballot."""
     choices = []
     for selection in contest.contest_selection:
         assert isinstance(selection, BallotMeasureSelection), \
@@ -236,8 +230,8 @@ def extract_ballot_measure_contest(contest: BallotMeasureContest, index):
     return result
 
 
-def gather_contests(ballot_style: BallotStyle, index):
-    """Extract all contest information needed for ballots."""
+def extract_contests(ballot_style: BallotStyle, index):
+    """Extract contest subset needed for ballots."""
     contests = {
         kind: [] for kind in ("candidate", "ballot_measure")
     }
@@ -252,3 +246,9 @@ def gather_contests(ballot_style: BallotStyle, index):
             # Ignore other contest types
             print(f"Skipping contest of type {contest.model__type}")
     return contests
+
+
+def extract_ballot_styles(election_report: ElectionReport, index):
+    """Extract all ballot styles."""
+    for ballot_style in index.by_type("BallotStyle"):
+        yield ballot_style

From 67d3933becdb13fdaca95a69e65510ae5eba1286 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Fri, 9 Sep 2022 16:44:36 -0700
Subject: [PATCH 30/48] Data model for ballot measure choices.

---
 src/electos/ballotmaker/data/models.py | 45 ++++++++++++++++
 tests/models/test_ballot_choices.py    | 72 ++++++++++++++++++++++++++
 2 files changed, 117 insertions(+)
 create mode 100644 src/electos/ballotmaker/data/models.py
 create mode 100644 tests/models/test_ballot_choices.py

diff --git a/src/electos/ballotmaker/data/models.py b/src/electos/ballotmaker/data/models.py
new file mode 100644
index 0000000..d8b8ceb
--- /dev/null
+++ b/src/electos/ballotmaker/data/models.py
@@ -0,0 +1,45 @@
+"""Ballot data models."""
+
+from dataclasses import dataclass
+from typing import List
+
+
+# --- Type assertions
+#
+# Python dataclasses don't check field types at runtime, but we want to avoid
+# errors.
+
+
+def _check_type(instance, field, type_):
+    """Raise 'TypeError' if 'instance.field' isn't of class 'type'."""
+    value = getattr(instance, field)
+    if not isinstance(value, type_):
+        raise TypeError(
+            f"Field '{field}' is not of type '{type_.__name__}': {value}"
+        )
+
+
+def _check_type_hint(instance, field, type_):
+    """Raise 'TypeError' if 'instance.field' isn't of type model 'type'."""
+    value = getattr(instance, field)
+    if not isinstance(value, type_):
+        raise TypeError(
+            f"Field '{field}' is not of type '{type_._name}': {value}"
+        )
+
+
+# --- Ballot contest data models
+
+
+@dataclass
+class BallotChoiceData:
+
+    """Data for ballot measure selections."""
+
+    id: str
+    choice: str
+
+
+    def __post_init__(self):
+        _check_type(self, "id", str)
+        _check_type(self, "choice", str)
diff --git a/tests/models/test_ballot_choices.py b/tests/models/test_ballot_choices.py
new file mode 100644
index 0000000..8a4179b
--- /dev/null
+++ b/tests/models/test_ballot_choices.py
@@ -0,0 +1,72 @@
+import pytest
+
+from pytest import raises
+from contextlib import nullcontext as raises_none
+
+from electos.ballotmaker.data.models import (
+    BallotChoiceData,
+)
+
+
+# Tests
+
+BALLOT_CHOICE_TESTS = [
+    (
+        {
+            "id": "ballot-measure-1--yes",
+            "choice": "yes",
+        },
+        raises_none(),
+    ),
+    # Missing id
+    (
+        {
+            "choice": "yes",
+        },
+        raises(TypeError, match = "required positional argument: 'id'"),
+    ),
+    # Missing choice
+    (
+        {
+            "id": "ballot-measure-1--yes",
+        },
+        raises(TypeError, match = "required positional argument: 'choice'"),
+    ),
+    # Empty object
+    (
+        {},
+        raises(TypeError, match = "2 required positional arguments: 'id' and 'choice'"),
+    ),
+    # id is not a string
+    (
+        {
+            "id": 1,
+            "choice": "yes",
+        },
+        raises(TypeError, match = "Field 'id' is not of type 'str'"),
+    ),
+    # choice is not a string
+    (
+        {
+            "id": "ballot-measure-1--yes",
+            "choice": [],
+        },
+        raises(TypeError, match = "Field 'choice' is not of type 'str'"),
+    ),
+]
+
+
+@pytest.mark.parametrize("data, raises", BALLOT_CHOICE_TESTS)
+def test_ballot_choice(data, raises):
+    with raises:
+        item = BallotChoiceData(**data)
+
+
+def test_ballot_choice_fields():
+    data = {
+        "id": "ballot-measure-1--yes",
+        "choice": "yes"
+    }
+    item = BallotChoiceData(**data)
+    assert item.id == data["id"]
+    assert item.choice == data["choice"]

From 9921c2d46f54729229423926468cdfd2cf6a7212 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Fri, 9 Sep 2022 17:30:11 -0700
Subject: [PATCH 31/48] Data model for ballot measure contests.

---
 src/electos/ballotmaker/data/models.py |  22 ++
 tests/models/test_ballot_contests.py   | 287 +++++++++++++++++++++++++
 2 files changed, 309 insertions(+)
 create mode 100644 tests/models/test_ballot_contests.py

diff --git a/src/electos/ballotmaker/data/models.py b/src/electos/ballotmaker/data/models.py
index d8b8ceb..472a6fd 100644
--- a/src/electos/ballotmaker/data/models.py
+++ b/src/electos/ballotmaker/data/models.py
@@ -43,3 +43,25 @@ class BallotChoiceData:
     def __post_init__(self):
         _check_type(self, "id", str)
         _check_type(self, "choice", str)
+
+
+@dataclass
+class BallotMeasureContestData:
+
+    """Data for ballot measure contests."""
+
+    id: str
+    type: str
+    title: str
+    district: str
+    text: str
+    choices: List[BallotChoiceData]
+
+    def __post_init__(self):
+        _check_type(self, "id", str)
+        _check_type(self, "type", str)
+        _check_type(self, "title", str)
+        _check_type(self, "district", str)
+        _check_type(self, "text", str)
+        _check_type_hint(self, "choices", List)
+        self.choices = [BallotChoiceData(**_) for _ in self.choices]
diff --git a/tests/models/test_ballot_contests.py b/tests/models/test_ballot_contests.py
new file mode 100644
index 0000000..48b1d74
--- /dev/null
+++ b/tests/models/test_ballot_contests.py
@@ -0,0 +1,287 @@
+import pytest
+
+from pytest import raises
+from contextlib import nullcontext as raises_none
+
+from electos.ballotmaker.data.models import (
+    BallotChoiceData,
+    BallotMeasureContestData,
+)
+
+
+# Tests
+
+BALLOT_MEASURE_CONTEST_TESTS = [
+    # Two choices
+    (
+        {
+            "id": "ballot-measure-1",
+            "type": "ballot-measure",
+            "title": "Ballot Measure #1",
+            "district": "Spacetown",
+            "text": "Ballot measure text",
+            "choices": [
+                {
+                    "id": "ballot-measure-1--yes",
+                    "choice": "yes",
+                },
+                {
+                    "id": "ballot-measure-1--no",
+                    "choice": "no",
+                },
+            ],
+        },
+        raises_none(),
+    ),
+    # Empty choices
+    pytest.param(
+        {
+            "id": "ballot-measure-1",
+            "type": "ballot-measure",
+            "title": "Ballot Measure #1",
+            "district": "Spacetown",
+            "text": "Ballot measure text",
+            "choices": [],
+        },
+        raises(ValueError, match = "Insufficient number of ballot choices"),
+        marks = pytest.mark.xfail(reason = "Empty 'choices' list"),
+    ),
+    # Only one choice
+    pytest.param(
+        {
+            "id": "ballot-measure-1",
+            "type": "ballot-measure",
+            "title": "Ballot Measure #1",
+            "district": "Spacetown",
+            "text": "Ballot measure text",
+            "choices": [
+                {
+                    "id": "ballot-measure-1--yes",
+                    "choice": "yes",
+                },
+            ],
+        },
+        raises(ValueError, match = "Insufficient number of ballot choices"),
+        marks = pytest.mark.xfail(reason = "Only one choice in 'choices' list"),
+    ),
+    # One choice, duplicated
+    pytest.param(
+        {
+            "id": "ballot-measure-1",
+            "type": "ballot-measure",
+            "title": "Ballot Measure #1",
+            "district": "Spacetown",
+            "text": "Ballot measure text",
+            "choices": [
+                {
+                    "id": "ballot-measure-1--yes",
+                    "choice": "yes",
+                },
+                {
+                    "id": "ballot-measure-1--yes",
+                    "choice": "yes",
+                },
+            ],
+        },
+        raises(ValueError, match = "Duplicate ballot choices"),
+        marks = pytest.mark.xfail(reason = "Duplicate ballot choices"),
+    ),
+    # Missing id
+    (
+        {
+            "type": "ballot-measure",
+            "title": "Ballot Measure #1",
+            "district": "Spacetown",
+            "text": "Ballot measure text",
+            "choices": [],
+        },
+        raises(TypeError, match = "required positional argument: 'id'"),
+    ),
+    # Missing type
+    (
+        {
+            "id": "ballot-measure-1",
+            "title": "Ballot Measure #1",
+            "district": "Spacetown",
+            "text": "Ballot measure text",
+            "choices": [],
+        },
+        raises(TypeError, match = "required positional argument: 'type'"),
+    ),
+    # Missing title
+    (
+        {
+            "type": "ballot-measure",
+            "id": "ballot-measure-1",
+            "district": "Spacetown",
+            "text": "Ballot measure text",
+            "choices": [],
+        },
+        raises(TypeError, match = "required positional argument: 'title'"),
+    ),
+    # Missing district
+    (
+        {
+            "id": "ballot-measure-1",
+            "type": "ballot-measure",
+            "title": "Ballot Measure #1",
+            "text": "Ballot measure text",
+            "choices": [],
+        },
+        raises(TypeError, match = "required positional argument: 'district'"),
+    ),
+    # Missing text
+    (
+        {
+            "id": "ballot-measure-1",
+            "type": "ballot-measure",
+            "title": "Ballot Measure #1",
+            "district": "Spacetown",
+            "choices": [],
+        },
+        raises(TypeError, match = "required positional argument: 'text'"),
+    ),
+    # Missing choices
+    (
+        {
+            "id": "ballot-measure-1",
+            "type": "ballot-measure",
+            "title": "Ballot Measure #1",
+            "district": "Spacetown",
+            "text": "Ballot measure text",
+        },
+        raises(TypeError, match = "required positional argument: 'choices'"),
+    ),
+    # Empty object
+    (
+        {},
+        raises(TypeError, match = "required positional arguments: 'id', 'type', 'title', 'district', 'text', and 'choices'"),
+    ),
+    # id is not a string
+    (
+        {
+            "id": 1,
+            "type": "ballot-measure",
+            "title": "Ballot Measure #1",
+            "district": "Spacetown",
+            "text": "Ballot measure text",
+            "choices": [],
+        },
+        raises(TypeError, match = "Field 'id' is not of type 'str'"),
+    ),
+    # type is not a string
+    (
+        {
+            "id": "ballot-measure-1",
+            "type": 1,
+            "title": "Ballot Measure #1",
+            "district": "Spacetown",
+            "text": "Ballot measure text",
+            "choices": [],
+        },
+        raises(TypeError, match = "Field 'type' is not of type 'str'"),
+    ),
+    # title is not a string
+    (
+        {
+            "id": "ballot-measure-1",
+            "type": "ballot-measure",
+            "title": 2,
+            "district": "Spacetown",
+            "text": "Ballot measure text",
+            "choices": [],
+        },
+        raises(TypeError, match = "Field 'title' is not of type 'str'"),
+    ),
+    # district is not a string
+    (
+        {
+            "id": "ballot-measure-1",
+            "type": "ballot-measure",
+            "title": "Ballot Measure #1",
+            "district": [],
+            "text": "Ballot measure text",
+            "choices": [],
+        },
+        raises(TypeError, match = "Field 'district' is not of type 'str'"),
+    ),
+    # text is not a string
+    (
+        {
+            "id": "ballot-measure-1",
+            "type": "ballot-measure",
+            "title": "Ballot Measure #1",
+            "district": "Spacetown",
+            "text": {},
+            "choices": [],
+        },
+        raises(TypeError, match = "Field 'text' is not of type 'str'"),
+    ),
+    # choices is not a list
+    (
+        {
+            "id": "ballot-measure-1",
+            "type": "ballot-measure",
+            "title": "Ballot Measure #1",
+            "district": "Spacetown",
+            "text": "Ballot measure text",
+            "choices": {},
+        },
+        raises(TypeError, match = "Field 'choices' is not of type 'List'"),
+    ),
+    # choice is not a dictionary that can convert to a BallotChoiceData
+    (
+        {
+            "id": "ballot-measure-1",
+            "type": "ballot-measure",
+            "title": "Ballot Measure #1",
+            "district": "Spacetown",
+            "text": "Ballot measure text",
+            "choices": [{}],
+        },
+        raises(TypeError, match = "missing 2 required positional arguments: 'id' and 'choice'"),
+    ),
+]
+
+
+@pytest.mark.parametrize("data, raises", BALLOT_MEASURE_CONTEST_TESTS)
+def test_ballot_measure_contest(data, raises):
+    with raises:
+        item = BallotMeasureContestData(**data)
+
+
+def test_ballot_measure_contest_fields():
+    data = {
+        "id": "ballot-measure-1",
+        "type": "ballot-measure",
+        "title": "Ballot Measure #1",
+        "district": "Spacetown",
+        "text": "Ballot measure text",
+        "choices": [
+            {
+                "id": "ballot-measure-1--yes",
+                "choice": "yes"
+            },
+            {
+                "id": "ballot-measure-1--no",
+                "choice": "no"
+            },
+        ]
+    }
+    item = BallotMeasureContestData(**data)
+    # Scalar fields match
+    assert item.id == data["id"]
+    assert item.type == data["type"]
+    assert item.title == data["title"]
+    assert item.district == data["district"]
+    assert item.text == data["text"]
+    # Not the same type: data model converts each party to an object
+    assert item.choices != data["choices"]
+    # Scalar field values
+    assert item.type == "ballot-measure"
+    # Lengths and fields are the same.
+    assert len(item.choices) == len(data["choices"])
+    assert all(isinstance(_, BallotChoiceData) for _ in item.choices)
+    for actual, expected in zip(item.choices, data["choices"]):
+        assert actual.id == expected["id"]
+        assert actual.choice == expected["choice"]

From 4e7283f2a6bb156f6764e949459052700048ce5d Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Fri, 9 Sep 2022 18:09:22 -0700
Subject: [PATCH 32/48] Data model for candidate parties.

---
 src/electos/ballotmaker/data/models.py | 16 ++++++
 tests/models/test_parties.py           | 72 ++++++++++++++++++++++++++
 2 files changed, 88 insertions(+)
 create mode 100644 tests/models/test_parties.py

diff --git a/src/electos/ballotmaker/data/models.py b/src/electos/ballotmaker/data/models.py
index 472a6fd..911713f 100644
--- a/src/electos/ballotmaker/data/models.py
+++ b/src/electos/ballotmaker/data/models.py
@@ -65,3 +65,19 @@ def __post_init__(self):
         _check_type(self, "text", str)
         _check_type_hint(self, "choices", List)
         self.choices = [BallotChoiceData(**_) for _ in self.choices]
+
+
+# --- Candidate contest data models
+
+
+@dataclass
+class PartyData:
+
+    """Data for parties candidates are in."""
+
+    name: str
+    abbreviation: str
+
+    def __post_init__(self):
+        _check_type(self, "name", str)
+        _check_type(self, "abbreviation", str)
diff --git a/tests/models/test_parties.py b/tests/models/test_parties.py
new file mode 100644
index 0000000..0cdaf6e
--- /dev/null
+++ b/tests/models/test_parties.py
@@ -0,0 +1,72 @@
+import pytest
+
+from pytest import raises
+from contextlib import nullcontext as raises_none
+
+from electos.ballotmaker.data.models import (
+    PartyData,
+)
+
+
+# Tests
+
+PARTY_TESTS = [
+    (
+        {
+            "name": "Un-Committed Party",
+            "abbreviation": "UCP",
+        },
+        raises_none(),
+    ),
+    # Missing name
+    (
+        {
+            "abbreviation": "UCP",
+        },
+        raises(TypeError, match = "required positional argument: 'name'"),
+    ),
+    # Missing abbreviation
+    (
+        {
+            "name": "Un-Committed Party",
+        },
+        raises(TypeError, match = "required positional argument: 'abbreviation'"),
+    ),
+    # Empty object
+    (
+        {},
+        raises(TypeError, match = "2 required positional arguments: 'name' and 'abbreviation'"),
+    ),
+    # name is not a string
+    (
+        {
+            "name": 1,
+            "abbreviation": "UCP",
+        },
+        raises(TypeError, match = "Field 'name' is not of type 'str'"),
+    ),
+    # abbreviation is not a string
+    (
+        {
+            "name": "Un-Committed Party",
+            "abbreviation": [],
+        },
+        raises(TypeError, match = "Field 'abbreviation' is not of type 'str'"),
+    ),
+]
+
+
+@pytest.mark.parametrize("data, raises", PARTY_TESTS)
+def test_party(data, raises):
+    with raises:
+        item = PartyData(**data)
+
+
+def test_party_fields():
+    data = {
+        "name": "Un-Committed Party",
+        "abbreviation": "UCP",
+    }
+    item = PartyData(**data)
+    assert item.name == data["name"]
+    assert item.abbreviation == data["abbreviation"]

From e0994191f959a70ae99b081e928e6a389c14b54c Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Fri, 9 Sep 2022 18:59:01 -0700
Subject: [PATCH 33/48] Data model for candidate choices.

---
 src/electos/ballotmaker/data/models.py |  28 +++
 tests/models/test_candidate_choices.py | 314 +++++++++++++++++++++++++
 2 files changed, 342 insertions(+)
 create mode 100644 tests/models/test_candidate_choices.py

diff --git a/src/electos/ballotmaker/data/models.py b/src/electos/ballotmaker/data/models.py
index 911713f..0105696 100644
--- a/src/electos/ballotmaker/data/models.py
+++ b/src/electos/ballotmaker/data/models.py
@@ -28,6 +28,15 @@ def _check_type_hint(instance, field, type_):
         )
 
 
+def _check_type_list(instance, field, type_):
+    """Raise 'TypeError' if 'instance.field' contents aren't of type 'type'."""
+    values = getattr(instance, field)
+    if not all(isinstance(value, type_) for value in values):
+        raise TypeError(
+            f"Values in field '{field}' are not all of type '{type_.__name__}': {values}"
+        )
+
+
 # --- Ballot contest data models
 
 
@@ -81,3 +90,22 @@ class PartyData:
     def __post_init__(self):
         _check_type(self, "name", str)
         _check_type(self, "abbreviation", str)
+
+
+@dataclass
+class CandidateChoiceData:
+
+    """Data for candidate contest selections."""
+
+    id: str
+    name: List[str]
+    party: List[PartyData]
+    is_write_in: bool
+
+    def __post_init__(self):
+        _check_type(self, "id", str)
+        _check_type_hint(self, "name", List)
+        _check_type_list(self, "name", str)
+        _check_type_hint(self, "party", List)
+        self.party = [PartyData(**_) for _ in self.party]
+        _check_type(self, "is_write_in", bool)
diff --git a/tests/models/test_candidate_choices.py b/tests/models/test_candidate_choices.py
new file mode 100644
index 0000000..a2cfddc
--- /dev/null
+++ b/tests/models/test_candidate_choices.py
@@ -0,0 +1,314 @@
+import pytest
+
+from pytest import raises
+from contextlib import nullcontext as raises_none
+
+from electos.ballotmaker.data.models import (
+    CandidateChoiceData,
+    PartyData,
+)
+
+
+# Tests
+
+CANDIDATE_CHOICE_TESTS = [
+    # Single candidate, no party
+    (
+        {
+            "id": "candidate-contest-1--candidate-1",
+            "name": [
+                "Anthony Alpha",
+            ],
+            "party": [],
+            "is_write_in": False,
+        },
+        raises_none(),
+    ),
+    # Multiple candidates, no party
+    (
+        {
+            "id": "candidate-contest-1--candidate-1",
+            "name": [
+                "Anthony Alpha",
+                "Betty Beta",
+            ],
+            "party": [],
+            "is_write_in": False,
+        },
+        raises_none(),
+    ),
+    # Multiple candidates, single party
+    (
+        {
+            "id": "candidate-contest-1--candidate-1",
+            "name": [
+                "Anthony Alpha",
+                "Betty Beta",
+            ],
+            "party": [
+                {
+                    "name": "Lepton Party",
+                    "abbreviation": "LEP",
+                },
+            ],
+            "is_write_in": False,
+        },
+        raises_none(),
+    ),
+    # Multiple candidates, same number of parties
+    (
+        {
+            "id": "candidate-contest-1--candidate-1",
+            "name": [
+                "Anthony Alpha",
+                "Elizabeth Epsilon",
+            ],
+            "party": [
+                {
+                    "name": "Lepton Party",
+                    "abbreviation": "LEP",
+                },
+                {
+                    "name": "Fermion Party",
+                    "abbreviation": "FRM",
+                },
+            ],
+            "is_write_in": False,
+        },
+        raises_none(),
+    ),
+    # Multiple candidates, differing number of parties
+    # Note: should have the same number of parties iff no. of candidates > 1
+    pytest.param(
+        {
+            "id": "candidate-contest-1--candidate-1",
+            "name": [
+                "Anthony Alpha",
+                "Elizabeth Epsilon",
+            ],
+            "party": [
+                {
+                    "name": "Lepton Party",
+                    "abbreviation": "LEP",
+                },
+                {
+                    "name": "Hadron Party",
+                    "abbreviation": "HAD",
+                },
+                {
+                    "name": "Fermion Party",
+                    "abbreviation": "FRM",
+                },
+            ],
+            "is_write_in": False,
+        },
+        raises(ValueError, match = "Counts of names and parties don't match"),
+        marks = pytest.mark.xfail(reason = "Mismatched counts of candidate names and parties"),
+    ),
+    # 'name' shouldn't be empty if 'is_write_in' is False
+    pytest.param(
+        {
+            "id": "candidate-contest-1--candidate-1",
+            "name": [],
+            "party": [],
+            "is_write_in": False,
+        },
+        raises(ValueError, match = "Candidate name is empty"),
+        marks = pytest.mark.xfail(reason = "Empty 'name' list"),
+    ),
+    # 'name' is empty if 'is_write_in' is True
+    (
+        {
+            "id": "candidate-contest-1--candidate-1",
+            "name": [],
+            "party": [],
+            "is_write_in": True,
+        },
+        raises_none(),
+    ),
+    # 'name' must be empty if 'is_write_in' is True
+    pytest.param(
+        {
+            "id": "candidate-contest-1--candidate-1",
+            "name": [
+                "Anthony Alpha",
+            ],
+            "party": [],
+            "is_write_in": True,
+        },
+        raises(ValueError, match = "Write-in cannot have a 'name'"),
+        marks = pytest.mark.xfail(reason = "Write-ins can't have candidate names"),
+    ),
+    # 'party' must be empty if 'is_write_in' is True
+    pytest.param(
+        {
+            "id": "candidate-contest-1--candidate-1",
+            "name": [],
+            "party": [
+                {
+                    "name": "Lepton Party",
+                    "abbreviation": "LEP",
+                },
+            ],
+            "is_write_in": True,
+        },
+        raises(ValueError, match = "Write-in cannot have a 'party'"),
+        marks = pytest.mark.xfail(reason = "Write-ins can't have candidate parties"),
+    ),
+    # Missing ID
+    (
+        {
+            "name": [
+                "Anthony Alpha",
+            ],
+            "party": [
+                {
+                    "name": "Lepton Party",
+                    "abbreviation": "LEP",
+                },
+            ],
+            "is_write_in": False,
+        },
+        raises(TypeError, match = "required positional argument: 'id'"),
+    ),
+    # Missing name
+    (
+        {
+            "id": "candidate-contest-1--candidate-1",
+            "party": [
+                {
+                    "name": "Lepton Party",
+                    "abbreviation": "LEP",
+                },
+            ],
+            "is_write_in": False,
+        },
+        raises(TypeError, match = "required positional argument: 'name'"),
+    ),
+    # Missing party
+    (
+        {
+            "id": "candidate-contest-1--candidate-1",
+            "name": [
+                "Anthony Alpha",
+            ],
+            "is_write_in": False,
+        },
+        raises(TypeError, match = "required positional argument: 'party'"),
+    ),
+    # Missing 'is write-in'
+    (
+        {
+            "id": "candidate-contest-1--candidate-1",
+            "name": [
+                "Anthony Alpha",
+            ],
+            "party": [
+                {
+                    "name": "Lepton Party",
+                    "abbreviation": "LEP",
+                },
+            ],
+        },
+        raises(TypeError, match = "required positional argument: 'is_write_in'"),
+    ),
+    # Empty object
+    (
+        {},
+        raises(TypeError, match = "required positional arguments: 'id', 'name', 'party', and 'is_write_in'"),
+    ),
+    # id is not a string
+    (
+        {
+            "id": 1,
+            "name": [
+                "Anthony Alpha",
+            ],
+            "party": [],
+            "is_write_in": False,
+        },
+        raises(TypeError, match = "Field 'id' is not of type 'str'"),
+    ),
+    # name is not a list
+    (
+        {
+            "id": "candidate-contest-1--candidate-1",
+            "name": "Anthony Alpha",
+            "party": [],
+            "is_write_in": False,
+        },
+        raises(TypeError, match = "Field 'name' is not of type 'List'"),
+    ),
+    # Contents of 'name' are not all strings
+    (
+        {
+            "id": "candidate-contest-1--candidate-1",
+            "name": [
+                "Anthony Alpha",
+                { "name": "Betty Beta" },
+            ],
+            "party": [],
+            "is_write_in": False,
+        },
+        raises(TypeError, match = "Values in field 'name' are not all of type 'str'"),
+    ),
+    # party is not a list
+    (
+        {
+            "id": "candidate-contest-1--candidate-1",
+            "name": [
+                "Anthony Alpha",
+            ],
+            "party": "",
+            "is_write_in": False,
+        },
+        raises(TypeError, match = "Field 'party' is not of type 'List'"),
+    ),
+    # 'is write-in' is not a boolean
+    (
+        {
+            "id": "candidate-contest-1--candidate-1",
+            "name": [
+                "Anthony Alpha",
+            ],
+            "party": [],
+            "is_write_in": None,
+        },
+        raises(TypeError, match = "Field 'is_write_in' is not of type 'bool'"),
+    ),
+]
+
+
+@pytest.mark.parametrize("data, raises", CANDIDATE_CHOICE_TESTS)
+def test_candidate_choice(data, raises):
+    with raises:
+        item = CandidateChoiceData(**data)
+
+
+def test_candidate_choice_fields():
+    data = {
+        "id": "candidate-contest-1--candidate-1",
+        "name": [
+            "Anthony Alpha",
+            "Betty Beta",
+        ],
+        "party": [
+            {
+                "name": "Lepton Party",
+                "abbreviation": "LEP",
+            },
+        ],
+        "is_write_in": False,
+    }
+    item = CandidateChoiceData(**data)
+    assert item.id == data["id"]
+    assert item.name == data["name"]
+    # Not the same type: data model converts each party to an object
+    assert item.party != data["party"]
+    # Lengths and fields are the same.
+    assert len(item.party) == len(data["party"])
+    assert all(isinstance(_, PartyData) for _ in item.party)
+    for actual, expected in zip(item.party, data["party"]):
+        assert actual.name == expected["name"]
+        assert actual.abbreviation == expected["abbreviation"]
+    assert item.is_write_in == data["is_write_in"]

From 5dc5ca8f66a957207792e95035e84a91a1fc68bc Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Fri, 9 Sep 2022 19:28:40 -0700
Subject: [PATCH 34/48] Data model for candidate contests.

---
 src/electos/ballotmaker/data/models.py  |  24 ++
 tests/models/test_candidate_contests.py | 523 ++++++++++++++++++++++++
 2 files changed, 547 insertions(+)
 create mode 100644 tests/models/test_candidate_contests.py

diff --git a/src/electos/ballotmaker/data/models.py b/src/electos/ballotmaker/data/models.py
index 0105696..12305c6 100644
--- a/src/electos/ballotmaker/data/models.py
+++ b/src/electos/ballotmaker/data/models.py
@@ -109,3 +109,27 @@ def __post_init__(self):
         _check_type_hint(self, "party", List)
         self.party = [PartyData(**_) for _ in self.party]
         _check_type(self, "is_write_in", bool)
+
+
+@dataclass
+class CandidateContestData:
+
+    """Data for candidate contests."""
+
+    id: str
+    type: str
+    title: str
+    district: str
+    vote_type: str
+    votes_allowed: str
+    candidates: List[CandidateChoiceData]
+
+    def __post_init__(self):
+        _check_type(self, "id", str)
+        _check_type(self, "type", str)
+        _check_type(self, "title", str)
+        _check_type(self, "district", str)
+        _check_type(self, "vote_type", str)
+        _check_type(self, "votes_allowed", int)
+        _check_type_hint(self, "candidates", List)
+        self.candidates = [CandidateChoiceData(**_) for _ in self.candidates]
diff --git a/tests/models/test_candidate_contests.py b/tests/models/test_candidate_contests.py
new file mode 100644
index 0000000..934dd6c
--- /dev/null
+++ b/tests/models/test_candidate_contests.py
@@ -0,0 +1,523 @@
+import pytest
+
+from pytest import raises
+from contextlib import nullcontext as raises_none
+
+from dataclasses import asdict
+
+from electos.ballotmaker.data.models import (
+    CandidateChoiceData,
+    CandidateContestData,
+)
+
+
+# Tests
+
+CANDIDATE_CONTEST_TESTS = [
+    # Single candidate
+    (
+        {
+            "id": "candidate-contest-1",
+            "type": "candidate",
+            "title": "Mayor of Orbit City",
+            "district": "Orbit City",
+            "vote_type": "plurality",
+            "votes_allowed": 1,
+            "candidates": [
+                {
+                    "id": "candidate-contest-1--candidate-1",
+                    "name": [
+                        "Spencer Cogswell",
+                    ],
+                    "party": [
+                        {
+                            "name": "The Hadron Party of Farallon",
+                            "abbreviation": "HAD",
+                        },
+                    ],
+                    "is_write_in": False,
+                },
+            ],
+        },
+        raises_none(),
+    ),
+    # Multiple candidates
+    (
+        {
+            "id": "candidate-contest-1",
+            "type": "candidate",
+            "title": "Mayor of Orbit City",
+            "district": "Orbit City",
+            "vote_type": "plurality",
+            "votes_allowed": 1,
+            "candidates": [
+                {
+                    "id": "candidate-contest-1--candidate-1",
+                    "name": [
+                        "Cosmo Spacely",
+                    ],
+                    "party": [
+                        {
+                            "name": "The Lepton Partyn",
+                            "abbreviation": "LEP",
+                        },
+                    ],
+                    "is_write_in": False,
+                },
+                {
+                    "id": "candidate-contest-1--candidate-2",
+                    "name": [
+                        "Spencer Cogswell",
+                    ],
+                    "party": [
+                        {
+                            "name": "The Hadron Party of Farallon",
+                            "abbreviation": "HAD",
+                        },
+                    ],
+                    "is_write_in": False,
+                },
+            ],
+        },
+        raises_none(),
+    ),
+    # Write-in
+    (
+        {
+            "id": "candidate-contest-1",
+            "type": "candidate",
+            "title": "Mayor of Orbit City",
+            "district": "Orbit City",
+            "vote_type": "plurality",
+            "votes_allowed": 1,
+            "candidates": [
+                {
+                    "id": "candidate-contest-1--write-in-1",
+                    "name": [],
+                    "party": [],
+                    "is_write_in": True,
+                },
+            ],
+        },
+        raises_none(),
+    ),
+    # Candidate + write-in
+    (
+        {
+            "id": "candidate-contest-1",
+            "type": "candidate",
+            "title": "Mayor of Orbit City",
+            "district": "Orbit City",
+            "vote_type": "plurality",
+            "votes_allowed": 1,
+            "candidates": [
+                {
+                    "id": "candidate-contest-1--candidate-1",
+                    "name": [
+                        "Spencer Cogswell",
+                    ],
+                    "party": [
+                        {
+                            "name": "The Hadron Party of Farallon",
+                            "abbreviation": "HAD",
+                        }
+                    ],
+                    "is_write_in": False,
+                },
+                {
+                    "id": "candidate-contest-1--write-in-1",
+                    "name": [],
+                    "party": [],
+                    "is_write_in": True,
+                },
+            ],
+        },
+        raises_none(),
+    ),
+    # Empty candidates
+    pytest.param(
+        {
+            "id": "candidate-contest-1",
+            "type": "candidate",
+            "title": "Mayor of Orbit City",
+            "district": "Orbit City",
+            "vote_type": "plurality",
+            "votes_allowed": 1,
+            "candidates": [],
+        },
+        raises(ValueError, match = "Insufficient number of candidates"),
+        marks = pytest.mark.xfail(reason = "Empty 'candidates' list"),
+    ),
+    # Single candidate, duplicated
+    pytest.param(
+        {
+            "id": "candidate-contest-1",
+            "type": "candidate",
+            "title": "Mayor of Orbit City",
+            "district": "Orbit City",
+            "vote_type": "plurality",
+            "votes_allowed": 1,
+            "candidates": [
+                {
+                    "id": "candidate-contest-1--candidate-1",
+                    "name": [
+                        "Spencer Cogswell",
+                    ],
+                    "party": [
+                        {
+                            "name": "The Hadron Party of Farallon",
+                            "abbreviation": "HAD",
+                        },
+                    ],
+                    "is_write_in": False,
+                },
+                {
+                    "id": "candidate-contest-1--candidate-1",
+                    "name": [
+                        "Spencer Cogswell",
+                    ],
+                    "party": [
+                        {
+                            "name": "The Hadron Party of Farallon",
+                            "abbreviation": "HAD",
+                        },
+                    ],
+                    "is_write_in": False,
+                },
+            ],
+        },
+        raises(ValueError, match = "Duplicate candidate"),
+        marks = pytest.mark.xfail(reason = "Duplicate candidate"),
+    ),
+    # Missing id
+    (
+        {
+            "type": "candidate",
+            "title": "Mayor of Orbit City",
+            "district": "Orbit City",
+            "vote_type": "plurality",
+            "votes_allowed": 1,
+            "candidates": [
+                {
+                    "id": "candidate-contest-1--candidate-1",
+                    "name": [
+                        "Spencer Cogswell",
+                    ],
+                    "party": [
+                        {
+                            "name": "The Hadron Party of Farallon",
+                            "abbreviation": "HAD"
+                        },
+                    ],
+                    "is_write_in": False,
+                },
+            ],
+        },
+        raises(TypeError, match = "required positional argument: 'id'"),
+    ),
+    # Missing type
+    (
+        {
+            "id": "candidate-contest-1",
+            "title": "Mayor of Orbit City",
+            "district": "Orbit City",
+            "vote_type": "plurality",
+            "votes_allowed": 1,
+            "candidates": [
+                {
+                    "id": "candidate-contest-1--candidate-1",
+                    "name": [
+                        "Spencer Cogswell",
+                    ],
+                    "party": [
+                        {
+                            "name": "The Hadron Party of Farallon",
+                            "abbreviation": "HAD"
+                        },
+                    ],
+                    "is_write_in": False,
+                },
+            ],
+        },
+        raises(TypeError, match = "required positional argument: 'type'"),
+    ),
+    # Missing title
+    (
+        {
+            "id": "candidate-contest-1",
+            "type": "candidate",
+            "district": "Orbit City",
+            "vote_type": "plurality",
+            "votes_allowed": 1,
+            "candidates": [
+                {
+                    "id": "candidate-contest-1--candidate-1",
+                    "name": [
+                        "Spencer Cogswell",
+                    ],
+                    "party": [
+                        {
+                            "name": "The Hadron Party of Farallon",
+                            "abbreviation": "HAD"
+                        },
+                    ],
+                    "is_write_in": False,
+                },
+            ],
+        },
+        raises(TypeError, match = "required positional argument: 'title'"),
+    ),
+    # Missing district
+    (
+        {
+            "id": "candidate-contest-1",
+            "type": "candidate",
+            "title": "Mayor of Orbit City",
+            "vote_type": "plurality",
+            "votes_allowed": 1,
+            "candidates": [
+                {
+                    "id": "candidate-contest-1--candidate-1",
+                    "name": [
+                        "Spencer Cogswell",
+                    ],
+                    "party": [
+                        {
+                            "name": "The Hadron Party of Farallon",
+                            "abbreviation": "HAD"
+                        },
+                    ],
+                    "is_write_in": False,
+                },
+            ],
+        },
+        raises(TypeError, match = "required positional argument: 'district'"),
+    ),
+    # Missing vote type
+    (
+        {
+            "id": "candidate-contest-1",
+            "type": "candidate",
+            "title": "Mayor of Orbit City",
+            "district": "Orbit City",
+            "votes_allowed": 1,
+            "candidates": [
+                {
+                    "id": "candidate-contest-1--candidate-1",
+                    "name": [
+                        "Spencer Cogswell",
+                    ],
+                    "party": [
+                        {
+                            "name": "The Hadron Party of Farallon",
+                            "abbreviation": "HAD"
+                        },
+                    ],
+                    "is_write_in": False,
+                },
+            ],
+        },
+        raises(TypeError, match = "required positional argument: 'vote_type'"),
+    ),
+    # Missing votes allowed
+    (
+        {
+            "id": "candidate-contest-1",
+            "type": "candidate",
+            "title": "Mayor of Orbit City",
+            "district": "Orbit City",
+            "vote_type": "plurality",
+            "candidates": [
+                {
+                    "id": "candidate-contest-1--candidate-1",
+                    "name": [
+                        "Spencer Cogswell",
+                    ],
+                    "party": [
+                        {
+                            "name": "The Hadron Party of Farallon",
+                            "abbreviation": "HAD"
+                        },
+                    ],
+                    "is_write_in": False,
+                },
+            ],
+        },
+        raises(TypeError, match = "required positional argument: 'votes_allowed'"),
+    ),
+    # Missing candidates
+    (
+        {
+            "id": "candidate-contest-1",
+            "type": "candidate",
+            "title": "Mayor of Orbit City",
+            "district": "Orbit City",
+            "vote_type": "plurality",
+            "votes_allowed": 1,
+        },
+        raises(TypeError, match = "required positional argument: 'candidates'"),
+    ),
+    # Empty object
+    (
+        {},
+        raises(TypeError, match = "required positional arguments: 'id', 'type', 'title', 'district', 'vote_type', 'votes_allowed', and 'candidates'"),
+    ),
+    # id is not a string
+    (
+        {
+            "id": 1,
+            "type": "candidate",
+            "title": "Mayor of Orbit City",
+            "district": "Orbit City",
+            "vote_type": "plurality",
+            "votes_allowed": 1,
+            "candidates": [],
+        },
+        raises(TypeError, match = "Field 'id' is not of type 'str'"),
+    ),
+    # type not a string
+    (
+        {
+            "id": "candidate-contest-1",
+            "type": 2,
+            "title": "Mayor of Orbit City",
+            "district": "Orbit City",
+            "vote_type": "plurality",
+            "votes_allowed": 1,
+            "candidates": [],
+        },
+        raises(TypeError, match = "Field 'type' is not of type 'str'"),
+    ),
+    # title is not a string
+    (
+        {
+            "id": "candidate-contest-1",
+            "type": "candidate",
+            "title": 3,
+            "district": "Orbit City",
+            "vote_type": "plurality",
+            "votes_allowed": 1,
+            "candidates": [],
+        },
+        raises(TypeError, match = "Field 'title' is not of type 'str'"),
+    ),
+    # district is not a string
+    (
+        {
+            "id": "candidate-contest-1",
+            "type": "candidate",
+            "title": "Mayor of Orbit City",
+            "district": 4,
+            "vote_type": "plurality",
+            "votes_allowed": 1,
+            "candidates": [],
+        },
+        raises(TypeError, match = "Field 'district' is not of type 'str'"),
+    ),
+    # "vote type" is not a string
+    (
+        {
+            "id": "candidate-contest-1",
+            "type": "candidate",
+            "title": "Mayor of Orbit City",
+            "district": "Orbit City",
+            "vote_type": 5,
+            "votes_allowed": 1,
+            "candidates": [],
+        },
+        raises(TypeError, match = "Field 'vote_type' is not of type 'str'"),
+    ),
+    # "votes allowed" not an integer
+    (
+        {
+            "id": "candidate-contest-1",
+            "type": "candidate",
+            "title": "Mayor of Orbit City",
+            "district": "Orbit City",
+            "vote_type": "plurality",
+            "votes_allowed": "one",
+            "candidates": [],
+        },
+        raises(TypeError, match = "Field 'votes_allowed' is not of type 'int'"),
+    ),
+    # candidates is not a list
+    (
+        {
+            "id": "candidate-contest-1",
+            "type": "candidate",
+            "title": "Mayor of Orbit City",
+            "district": "Orbit City",
+            "vote_type": "plurality",
+            "votes_allowed": 1,
+            "candidates": {},
+        },
+        raises(TypeError, match = "Field 'candidates' is not of type 'List'"),
+    ),
+    # candidate is not a dictionary that can convert to a CandidateChoiceData
+    (
+        {
+            "id": "candidate-contest-1",
+            "type": "candidate",
+            "title": "Mayor of Orbit City",
+            "district": "Orbit City",
+            "vote_type": "plurality",
+            "votes_allowed": 1,
+            "candidates": [{}],
+        },
+        raises(TypeError, match = "missing 4 required positional arguments: 'id', 'name', 'party', and 'is_write_in'"),
+    ),
+]
+
+
+@pytest.mark.parametrize("data, raises", CANDIDATE_CONTEST_TESTS)
+def test_candidate_contest(data, raises):
+    with raises:
+        item = CandidateContestData(**data)
+
+
+def test_candidate_contest_fields():
+    data = {
+        "id": "candidate-contest-1",
+        "type": "candidate",
+        "title": "Mayor of Orbit City",
+        "district": "Orbit City",
+        "vote_type": "plurality",
+        "votes_allowed": 1,
+        "candidates": [
+            {
+                "id": "candidate-contest-1--candidate-1",
+                "name": [
+                    "Spencer Cogswell",
+                ],
+                "party": [
+                    {
+                        "name": "The Hadron Party of Farallon",
+                        "abbreviation": "HAD",
+                    }
+                ],
+                "is_write_in": False,
+            },
+            {
+                "id": "candidate-contest-1--write-in-1",
+                "name": [],
+                "party": [],
+                "is_write_in": True,
+            },
+        ],
+    }
+    item = CandidateContestData(**data)
+    # Scalar fields match
+    assert item.id == data["id"]
+    assert item.type == data["type"]
+    assert item.title == data["title"]
+    assert item.vote_type == data["vote_type"]
+    assert item.votes_allowed == data["votes_allowed"]
+    assert item.district == data["district"]
+    # Not the same type: data model converts each party to an object
+    assert item.candidates != data["candidates"]
+    # Lengths and fields are the same.
+    assert len(item.candidates) == len(data["candidates"])
+    assert all(isinstance(_, CandidateChoiceData) for _ in item.candidates)
+    for actual, expected in zip(item.candidates, data["candidates"]):
+        actual = asdict(actual)
+        assert actual == expected

From a112e03f8e143c50273331bcd152a1d88c8c9e61 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Mon, 12 Sep 2022 09:09:48 -0700
Subject: [PATCH 35/48] Simplify test assertions on nested data by comparing
 dictionaries.

---
 tests/models/test_ballot_contests.py   | 6 ++++--
 tests/models/test_candidate_choices.py | 6 ++++--
 2 files changed, 8 insertions(+), 4 deletions(-)

diff --git a/tests/models/test_ballot_contests.py b/tests/models/test_ballot_contests.py
index 48b1d74..a46d6e3 100644
--- a/tests/models/test_ballot_contests.py
+++ b/tests/models/test_ballot_contests.py
@@ -3,6 +3,8 @@
 from pytest import raises
 from contextlib import nullcontext as raises_none
 
+from dataclasses import asdict
+
 from electos.ballotmaker.data.models import (
     BallotChoiceData,
     BallotMeasureContestData,
@@ -283,5 +285,5 @@ def test_ballot_measure_contest_fields():
     assert len(item.choices) == len(data["choices"])
     assert all(isinstance(_, BallotChoiceData) for _ in item.choices)
     for actual, expected in zip(item.choices, data["choices"]):
-        assert actual.id == expected["id"]
-        assert actual.choice == expected["choice"]
+        actual = asdict(actual)
+        assert actual == expected
diff --git a/tests/models/test_candidate_choices.py b/tests/models/test_candidate_choices.py
index a2cfddc..e965961 100644
--- a/tests/models/test_candidate_choices.py
+++ b/tests/models/test_candidate_choices.py
@@ -3,6 +3,8 @@
 from pytest import raises
 from contextlib import nullcontext as raises_none
 
+from dataclasses import asdict
+
 from electos.ballotmaker.data.models import (
     CandidateChoiceData,
     PartyData,
@@ -309,6 +311,6 @@ def test_candidate_choice_fields():
     assert len(item.party) == len(data["party"])
     assert all(isinstance(_, PartyData) for _ in item.party)
     for actual, expected in zip(item.party, data["party"]):
-        assert actual.name == expected["name"]
-        assert actual.abbreviation == expected["abbreviation"]
+        actual = asdict(actual)
+        assert actual == expected
     assert item.is_write_in == data["is_write_in"]

From 78738336603966ac39e50456bd34b944d35d87e3 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Mon, 12 Sep 2022 09:46:36 -0700
Subject: [PATCH 36/48] Data model for ballot styles.

---
 src/electos/ballotmaker/data/models.py |  39 +-
 tests/models/test_ballot_style.py      | 491 +++++++++++++++++++++++++
 2 files changed, 529 insertions(+), 1 deletion(-)
 create mode 100644 tests/models/test_ballot_style.py

diff --git a/src/electos/ballotmaker/data/models.py b/src/electos/ballotmaker/data/models.py
index 12305c6..83b31e4 100644
--- a/src/electos/ballotmaker/data/models.py
+++ b/src/electos/ballotmaker/data/models.py
@@ -1,7 +1,7 @@
 """Ballot data models."""
 
 from dataclasses import dataclass
-from typing import List
+from typing import List, Union
 
 
 # --- Type assertions
@@ -133,3 +133,40 @@ def __post_init__(self):
         _check_type(self, "votes_allowed", int)
         _check_type_hint(self, "candidates", List)
         self.candidates = [CandidateChoiceData(**_) for _ in self.candidates]
+
+
+@dataclass
+class BallotStyleData:
+
+    """Date for contests associated with a ballot style."""
+
+    BALLOT_MEASURE = "ballot measure"
+    CANDIDATE = "candidate"
+
+    # Note: Don't use separate fields for the types of contests.
+    # There's no guarantee the types will be clearly separated in an EDF.
+    # (The NIST SP-1500-100 JSON Schema uses unions too.)
+
+    id: str
+    scopes: List[str]
+    contests: List[Union[BallotMeasureContestData, CandidateContestData]]
+
+    def __post_init__(self):
+        _check_type(self, "id", str)
+        _check_type_hint(self, "scopes", List)
+        _check_type_list(self, "scopes", str)
+        _check_type_hint(self, "contests", List)
+        contests = []
+        for contest in self.contests:
+            if not isinstance(contest, dict):
+                raise TypeError(f"Contest is not a dictionary: '{contest}'")
+            if "type" not in contest:
+                raise KeyError(f"Contest has no 'type' field: '{contest}'")
+            if contest["type"] not in (self.BALLOT_MEASURE, self.CANDIDATE):
+                raise ValueError(f"Unhandled contest type: '{contest['type']}'")
+            if contest["type"] == self.BALLOT_MEASURE:
+                contest = BallotMeasureContestData(**contest)
+            elif contest["type"] == self.CANDIDATE:
+                contest = CandidateContestData(**contest)
+            contests.append(contest)
+        self.contests = contests
diff --git a/tests/models/test_ballot_style.py b/tests/models/test_ballot_style.py
new file mode 100644
index 0000000..eac2e0b
--- /dev/null
+++ b/tests/models/test_ballot_style.py
@@ -0,0 +1,491 @@
+import pytest
+
+from pytest import raises
+from contextlib import nullcontext as raises_none
+
+from dataclasses import asdict
+
+from electos.ballotmaker.data.models import (
+    BallotMeasureContestData,
+    BallotStyleData,
+    CandidateContestData,
+)
+
+
+# Tests
+
+BALLOT_STYLE_TESTS = [
+    # Single candidate contest
+    (
+        {
+            "id": "precinct_2_spacetown",
+            "scopes": [
+                "spacetown-precinct",
+            ],
+            "contests": [
+                {
+                    "id": "candidate-contest-1",
+                    "type": "candidate",
+                    "title": "Mayor of Orbit City",
+                    "district": "Orbit City",
+                    "vote_type": "plurality",
+                    "votes_allowed": 1,
+                    "candidates": [
+                        {
+                            "id": "candidate-contest-1--candidate-1",
+                            "name": [
+                                "Spencer Cogswell"
+                            ],
+                            "party": [
+                                {
+                                    "name": "The Hadron Party of Farallon",
+                                    "abbreviation": "HAD"
+                                },
+                            ],
+                            "is_write_in": False,
+                        },
+                    ],
+                },
+            ],
+        },
+        raises_none(),
+    ),
+    # Single ballot measure contest
+    (
+        {
+            "id": "precinct_2_spacetown",
+            "scopes": [
+                "spacetown-precinct",
+            ],
+            "contests": [
+                {
+                    "id": "ballot-measure-1",
+                    "type": "ballot measure",
+                    "title": "Ballot Measure #1",
+                    "district": "Spacetown",
+                    "text": "Ballot measure text",
+                    "choices": [
+                        {
+                            "id": "ballot-measure-1--yes",
+                            "choice": "yes",
+                        },
+                        {
+                            "id": "ballot-measure-1--no",
+                            "choice": "no",
+                        },
+                    ],
+                },
+            ],
+        },
+        raises_none(),
+    ),
+    # Multiple contests, conventional order
+    (
+        {
+            "id": "precinct_2_spacetown",
+            "scopes": [
+                "spacetown-precinct",
+            ],
+            "contests": [
+                {
+                    "id": "candidate-contest-1",
+                    "type": "candidate",
+                    "title": "Mayor of Orbit City",
+                    "district": "Orbit City",
+                    "vote_type": "plurality",
+                    "votes_allowed": 1,
+                    "candidates": [
+                        {
+                            "id": "candidate-contest-1--candidate-1",
+                            "name": [
+                                "Spencer Cogswell"
+                            ],
+                            "party": [
+                                {
+                                    "name": "The Hadron Party of Farallon",
+                                    "abbreviation": "HAD"
+                                },
+                            ],
+                            "is_write_in": False,
+                        },
+                    ],
+                },
+                {
+                    "id": "candidate-contest-2",
+                    "type": "candidate",
+                    "title": "President of the United States",
+                    "district": "United States of America",
+                    "vote_type": "plurality",
+                    "votes_allowed": 1,
+                    "candidates": [
+                        {
+                            "id": "candidate-contest-2--candidate-1",
+                            "name": [
+                                "Anthony Alpha",
+                                "Betty Beta"
+                            ],
+                            "party": [
+                                {
+                                    "name": "The Lepton Party",
+                                    "abbreviation": "LEP"
+                                }
+                            ],
+                            "is_write_in": False,
+                        },
+                        {
+                            "id": "candidate-contest-2--candidate-2",
+                            "name": [
+                                "Gloria Gamma",
+                                "David Delta"
+                            ],
+                            "party": [
+                                {
+                                    "name": "The Hadron Party of Farallon",
+                                    "abbreviation": "HAD"
+                                }
+                            ],
+                            "is_write_in": False,
+                        },
+                    ],
+                },
+                {
+                    "id": "ballot-measure-1",
+                    "type": "ballot measure",
+                    "title": "Ballot Measure #1",
+                    "district": "Spacetown",
+                    "text": "Ballot measure text",
+                    "choices": [
+                        {
+                            "id": "ballot-measure-1--yes",
+                            "choice": "yes",
+                        },
+                        {
+                            "id": "ballot-measure-1--no",
+                            "choice": "no",
+                        },
+                    ],
+                },
+            ],
+        },
+        raises_none(),
+    ),
+    # Multiple contests, candidates and write-ins intermingled
+    (
+        {
+            "id": "precinct_2_spacetown",
+            "scopes": [
+                "spacetown-precinct",
+            ],
+            "contests": [
+                {
+                    "id": "candidate-contest-1",
+                    "type": "candidate",
+                    "title": "Mayor of Orbit City",
+                    "district": "Orbit City",
+                    "vote_type": "plurality",
+                    "votes_allowed": 1,
+                    "candidates": [
+                        {
+                            "id": "candidate-contest-1--candidate-1",
+                            "name": [
+                                "Spencer Cogswell"
+                            ],
+                            "party": [
+                                {
+                                    "name": "The Hadron Party of Farallon",
+                                    "abbreviation": "HAD"
+                                },
+                            ],
+                            "is_write_in": False,
+                        },
+                    ],
+                },
+                {
+                    "id": "ballot-measure-1",
+                    "type": "ballot measure",
+                    "title": "Ballot Measure #1",
+                    "district": "Spacetown",
+                    "text": "Ballot measure text",
+                    "choices": [
+                        {
+                            "id": "ballot-measure-1--yes",
+                            "choice": "yes",
+                        },
+                        {
+                            "id": "ballot-measure-1--no",
+                            "choice": "no",
+                        },
+                    ],
+                },
+                {
+                    "id": "candidate-contest-2",
+                    "type": "candidate",
+                    "title": "President of the United States",
+                    "district": "United States of America",
+                    "vote_type": "plurality",
+                    "votes_allowed": 1,
+                    "candidates": [
+                        {
+                            "id": "candidate-contest-2--candidate-1",
+                            "name": [
+                                "Anthony Alpha",
+                                "Betty Beta"
+                            ],
+                            "party": [
+                                {
+                                    "name": "The Lepton Party",
+                                    "abbreviation": "LEP"
+                                }
+                            ],
+                            "is_write_in": False,
+                        },
+                        {
+                            "id": "candidate-contest-2--candidate-2",
+                            "name": [
+                                "Gloria Gamma",
+                                "David Delta"
+                            ],
+                            "party": [
+                                {
+                                    "name": "The Hadron Party of Farallon",
+                                    "abbreviation": "HAD"
+                                }
+                            ],
+                            "is_write_in": False,
+                        },
+                    ],
+                },
+            ],
+        },
+        raises_none(),
+    ),
+    # Empty contests
+    pytest.param(
+        {
+            "id": "precinct_2_spacetown",
+            "scopes": [
+                "spacetown-precinct",
+            ],
+            "contests": [],
+        },
+        raises(ValueError, match = "'contests' cannot be empty"),
+        marks = pytest.mark.xfail(reason = "Empty 'contest' list"),
+    ),
+    # Missing id
+    (
+        {
+            "scopes": [
+                "spacetown-precinct",
+            ],
+            "contests": [
+                {
+                    "id": "candidate-contest-1",
+                    "type": "candidate",
+                    "title": "Mayor of Orbit City",
+                    "district": "Orbit City",
+                    "vote_type": "plurality",
+                    "votes_allowed": 1,
+                    "candidates": [
+                        {
+                            "id": "candidate-contest-1--candidate-1",
+                            "name": [
+                                "Spencer Cogswell"
+                            ],
+                            "party": [
+                                {
+                                    "name": "The Hadron Party of Farallon",
+                                    "abbreviation": "HAD"
+                                },
+                            ],
+                            "is_write_in": False,
+                        },
+                    ],
+                },
+            ],
+        },
+        raises(TypeError, match = "1 required positional argument: 'id'"),
+    ),
+    # Missing scopes
+    (
+        {
+            "id": "precinct_2_spacetown",
+            "contests": [
+                {
+                    "id": "candidate-contest-1",
+                    "type": "candidate",
+                    "title": "Mayor of Orbit City",
+                    "district": "Orbit City",
+                    "vote_type": "plurality",
+                    "votes_allowed": 1,
+                    "candidates": [
+                        {
+                            "id": "candidate-contest-1--candidate-1",
+                            "name": [
+                                "Spencer Cogswell"
+                            ],
+                            "party": [
+                                {
+                                    "name": "The Hadron Party of Farallon",
+                                    "abbreviation": "HAD"
+                                },
+                            ],
+                            "is_write_in": False,
+                        },
+                    ],
+                },
+            ],
+        },
+        raises(TypeError, match = "1 required positional argument: 'scopes'"),
+    ),
+    # Missing contests
+    (
+        {
+            "id": "precinct_2_spacetown",
+            "scopes": [
+                "spacetown-precinct",
+            ],
+        },
+        raises(TypeError, match = "1 required positional argument: 'contests'"),
+    ),
+    # Empty object
+    (
+        {},
+        raises(TypeError, match = "3 required positional arguments: 'id', 'scopes', and 'contests'"),
+    ),
+    # contests is not a list
+    (
+        {
+            "id": "precinct_2_spacetown",
+            "scopes": [
+                "spacetown-precinct",
+            ],
+            "contests": {},
+        },
+        raises(TypeError, match = "Field 'contests' is not of type 'List'"),
+    ),
+    # Contest is not a dict
+    (
+        {
+            "id": "precinct_2_spacetown",
+            "scopes": [
+                "spacetown-precinct",
+            ],
+            "contests": [[]],
+        },
+        raises(TypeError, match = "Contest is not a dictionary"),
+    ),
+    # Contest has no type
+    (
+        {
+            "id": "precinct_2_spacetown",
+            "scopes": [
+                "spacetown-precinct",
+            ],
+            "contests": [{}],
+        },
+        raises(KeyError, match = "Contest has no 'type' field"),
+    ),
+    # Unhandled contest type
+    (
+        {
+            "id": "precinct_2_spacetown",
+            "scopes": [
+                "spacetown-precinct",
+            ],
+            "contests": [{ "type": "" }],
+        },
+        raises(ValueError, match = "Unhandled contest type: ''"),
+    ),
+    # Ballot measure contest
+    (
+        {
+            "id": "precinct_2_spacetown",
+            "scopes": [
+                "spacetown-precinct",
+            ],
+            "contests": [{ "type": BallotStyleData.BALLOT_MEASURE }],
+        },
+        raises(TypeError, match = "missing 5 required positional arguments: 'id', 'title', 'district', 'text', and 'choices'"),
+    ),
+    # Candidate contest
+    (
+        {
+            "id": "precinct_2_spacetown",
+            "scopes": [
+                "spacetown-precinct",
+            ],
+            "contests": [{ "type": BallotStyleData.CANDIDATE }],
+        },
+        raises(TypeError, match = "missing 6 required positional arguments: 'id', 'title', 'district', 'vote_type', 'votes_allowed', and 'candidates'"),
+    ),
+]
+
+
+@pytest.mark.parametrize("data, raises", BALLOT_STYLE_TESTS)
+def test_ballot_style(data, raises):
+    with raises:
+        item = BallotStyleData(**data)
+
+
+def test_ballot_style_fields():
+    data = {
+        "id": "precinct_2_spacetown",
+        "scopes": [
+            "spacetown-precinct",
+        ],
+        "contests": [
+            {
+                "id": "candidate-contest-1",
+                "type": "candidate",
+                "title": "Mayor of Orbit City",
+                "district": "Orbit City",
+                "vote_type": "plurality",
+                "votes_allowed": 1,
+                "candidates": [
+                    {
+                        "id": "candidate-contest-1--candidate-1",
+                        "name": [
+                            "Spencer Cogswell"
+                        ],
+                        "party": [
+                            {
+                                "name": "The Hadron Party of Farallon",
+                                "abbreviation": "HAD"
+                            },
+                        ],
+                        "is_write_in": False,
+                    },
+                ],
+            },
+            {
+                "id": "ballot-measure-1",
+                "type": "ballot measure",
+                "title": "Ballot Measure #1",
+                "district": "Spacetown",
+                "text": "Ballot measure text",
+                "choices": [
+                    {
+                        "id": "ballot-measure-1--yes",
+                        "choice": "yes",
+                    },
+                    {
+                        "id": "ballot-measure-1--no",
+                        "choice": "no",
+                    },
+                ],
+            },
+        ],
+    }
+    item = BallotStyleData(**data)
+    assert item.id == data["id"]
+    assert item.scopes == data["scopes"]
+    # Not the same type: data model converts each party to an object
+    assert item.contests != data["contests"]
+    # Lengths are the same.
+    assert len(item.contests) == len(data["contests"])
+    # Types are mixed
+    assert isinstance(item.contests[0], CandidateContestData)
+    assert isinstance(item.contests[1], BallotMeasureContestData)
+    for actual, expected in zip(item.contests, data["contests"]):
+        actual = asdict(actual)
+        assert actual == expected

From 3621f3268a5d0f3820b995597a98e82b65ffb722 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Mon, 12 Sep 2022 10:00:42 -0700
Subject: [PATCH 37/48] Data model for elections.

---
 src/electos/ballotmaker/data/models.py |  24 ++
 tests/models/test_elections.py         | 371 +++++++++++++++++++++++++
 2 files changed, 395 insertions(+)
 create mode 100644 tests/models/test_elections.py

diff --git a/src/electos/ballotmaker/data/models.py b/src/electos/ballotmaker/data/models.py
index 83b31e4..20c32f0 100644
--- a/src/electos/ballotmaker/data/models.py
+++ b/src/electos/ballotmaker/data/models.py
@@ -170,3 +170,27 @@ def __post_init__(self):
                 contest = CandidateContestData(**contest)
             contests.append(contest)
         self.contests = contests
+
+
+@dataclass
+class ElectionData:
+
+    """Data for elections."""
+
+    # Dates are not 'datetime' for simplicity and because it's assumed the date
+    # is formatter correctly. That can be changed.
+
+    name: str
+    type: str
+    start_date: str
+    end_date: str
+    ballot_styles: List[BallotStyleData]
+
+
+    def __post_init__(self):
+        _check_type(self, "name", str)
+        _check_type(self, "type", str)
+        _check_type(self, "start_date", str)
+        _check_type(self, "end_date", str)
+        _check_type_hint(self, "ballot_styles", List)
+        self.ballot_styles = [BallotStyleData(**_) for _ in self.ballot_styles]
diff --git a/tests/models/test_elections.py b/tests/models/test_elections.py
new file mode 100644
index 0000000..7356a3a
--- /dev/null
+++ b/tests/models/test_elections.py
@@ -0,0 +1,371 @@
+import pytest
+
+from pytest import raises
+from contextlib import nullcontext as raises_none
+
+from dataclasses import asdict
+
+from electos.ballotmaker.data.models import (
+    BallotStyleData,
+    ElectionData,
+)
+
+
+# Tests
+
+ELECTION_TESTS = [
+    (
+        {
+            "name": "General Election",
+            "type": "general",
+            "start_date": "2024-11-05",
+            "end_date": "2024-11-05",
+            "ballot_styles": [
+                {
+                    "id": "precinct_2_spacetown",
+                    "scopes": [
+                        "spacetown-precinct",
+                    ],
+                    "contests": [
+                        {
+                            "id": "candidate-contest-orbit-city-mayor",
+                            "title": "Mayor of Orbit City",
+                            "type": "candidate",
+                            "vote_type": "plurality",
+                            "votes_allowed": 1,
+                            "district": "Orbit City",
+                            "candidates": [
+                                {
+                                    "id": "candidate-choice-1",
+                                    "name": [
+                                        "Spencer Cogswell"
+                                    ],
+                                    "party": [
+                                        {
+                                            "name": "The Hadron Party of Farallon",
+                                            "abbreviation": "HAD"
+                                        },
+                                    ],
+                                    "is_write_in": False,
+                                },
+                            ],
+                        },
+                    ],
+                },
+            ],
+        },
+        raises_none()
+    ),
+    # Empty ballot styles
+    pytest.param(
+        {
+            "name": "General Election",
+            "start_date": "2024-11-05",
+            "end_date": "2024-11-05",
+            "type": "general",
+            "ballot_styles": [],
+        },
+        raises(ValueError, match = "Ballot styles are empty"),
+        marks = pytest.mark.xfail(reason = "Empty 'ballot_styles' list")
+    ),
+    # Missing name
+    (
+        {
+            "start_date": "2024-11-05",
+            "end_date": "2024-11-05",
+            "type": "general",
+            "ballot_styles": [
+                {
+                    "id": "precinct_2_spacetown",
+                    "scopes": [
+                        "spacetown-precinct",
+                    ],
+                    "contests": [
+                        {
+                            "id": "candidate-contest-orbit-city-mayor",
+                            "title": "Mayor of Orbit City",
+                            "type": "candidate",
+                            "vote_type": "plurality",
+                            "votes_allowed": 1,
+                            "district": "Orbit City",
+                            "candidates": [
+                                {
+                                    "id": "candidate-choice-1",
+                                    "name": [
+                                        "Spencer Cogswell"
+                                    ],
+                                    "party": [
+                                        {
+                                            "name": "The Hadron Party of Farallon",
+                                            "abbreviation": "HAD"
+                                        },
+                                    ],
+                                    "is_write_in": False,
+                                },
+                            ],
+                        },
+                    ],
+                },
+            ],
+        },
+        raises(TypeError, match = "required positional argument: 'name'")
+    ),
+    # Missing type
+    (
+        {
+            "name": "General Election",
+            "start_date": "2024-11-05",
+            "end_date": "2024-11-05",
+            "ballot_styles": [
+                {
+                    "id": "precinct_2_spacetown",
+                    "scopes": [
+                        "spacetown-precinct",
+                    ],
+                    "contests": [
+                        {
+                            "id": "candidate-contest-orbit-city-mayor",
+                            "title": "Mayor of Orbit City",
+                            "type": "candidate",
+                            "vote_type": "plurality",
+                            "votes_allowed": 1,
+                            "district": "Orbit City",
+                            "candidates": [
+                                {
+                                    "id": "candidate-choice-1",
+                                    "name": [
+                                        "Spencer Cogswell"
+                                    ],
+                                    "party": [
+                                        {
+                                            "name": "The Hadron Party of Farallon",
+                                            "abbreviation": "HAD"
+                                        },
+                                    ],
+                                    "is_write_in": False,
+                                },
+                            ],
+                        },
+                    ],
+                },
+            ],
+        },
+        raises(TypeError, match = "required positional argument: 'type'")
+    ),
+    # Missing start date
+    (
+        {
+            "name": "General Election",
+            "end_date": "2024-11-05",
+            "type": "general",
+            "ballot_styles": [
+                {
+                    "id": "precinct_2_spacetown",
+                    "scopes": [
+                        "spacetown-precinct",
+                    ],
+                    "contests": [
+                        {
+                            "id": "candidate-contest-orbit-city-mayor",
+                            "title": "Mayor of Orbit City",
+                            "type": "candidate",
+                            "vote_type": "plurality",
+                            "votes_allowed": 1,
+                            "district": "Orbit City",
+                            "candidates": [
+                                {
+                                    "id": "candidate-choice-1",
+                                    "name": [
+                                        "Spencer Cogswell"
+                                    ],
+                                    "party": [
+                                        {
+                                            "name": "The Hadron Party of Farallon",
+                                            "abbreviation": "HAD"
+                                        },
+                                    ],
+                                    "is_write_in": False,
+                                },
+                            ],
+                        },
+                    ],
+                },
+            ],
+        },
+        raises(TypeError, match = "required positional argument: 'start_date'")
+    ),
+    # Missing end date
+    (
+        {
+            "name": "General Election",
+            "start_date": "2024-11-05",
+            "type": "general",
+            "ballot_styles": [
+                {
+                    "id": "precinct_2_spacetown",
+                    "scopes": [
+                        "spacetown-precinct",
+                    ],
+                    "contests": [
+                        {
+                            "id": "candidate-contest-orbit-city-mayor",
+                            "title": "Mayor of Orbit City",
+                            "type": "candidate",
+                            "vote_type": "plurality",
+                            "votes_allowed": 1,
+                            "district": "Orbit City",
+                            "candidates": [
+                                {
+                                    "id": "candidate-choice-1",
+                                    "name": [
+                                        "Spencer Cogswell"
+                                    ],
+                                    "party": [
+                                        {
+                                            "name": "The Hadron Party of Farallon",
+                                            "abbreviation": "HAD"
+                                        },
+                                    ],
+                                    "is_write_in": False,
+                                },
+                            ],
+                        },
+                    ],
+                },
+            ],
+        },
+        raises(TypeError, match = "required positional argument: 'end_date'")
+    ),
+    # Missing ballot styles
+    (
+        {
+            "name": "General Election",
+            "start_date": "2024-11-05",
+            "end_date": "2024-11-05",
+            "type": "general",
+        },
+        raises(TypeError, match = "required positional argument: 'ballot_styles'")
+    ),
+    # Empty object
+    (
+        {},
+        raises(TypeError, match = "5 required positional arguments: 'name', 'type', 'start_date', 'end_date', and 'ballot_styles'")
+    ),
+    # name is not a string
+    (
+        {
+            "name": 1,
+            "type": "general",
+            "start_date": "2024-11-05",
+            "end_date": "2024-11-05",
+            "ballot_styles": [],
+        },
+        raises(TypeError, match = "Field 'name' is not of type 'str'"),
+    ),
+    # type is not a string
+    (
+        {
+            "name": "General Election",
+            "type": 2,
+            "start_date": "2024-11-05",
+            "end_date": "2024-11-05",
+            "ballot_styles": [],
+        },
+        raises(TypeError, match = "Field 'type' is not of type 'str'"),
+    ),
+    # 'start date' is not a string
+    (
+        {
+            "name": "General Election",
+            "type": "general",
+            "start_date": 3,
+            "end_date": "2024-11-05",
+            "ballot_styles": [],
+        },
+        raises(TypeError, match = "Field 'start_date' is not of type 'str'"),
+    ),
+    # 'end date' is not a string
+    (
+        {
+            "name": "General Election",
+            "type": "general",
+            "start_date": "2024-11-05",
+            "end_date": 4,
+            "ballot_styles": [],
+        },
+        raises(TypeError, match = "Field 'end_date' is not of type 'str'"),
+    ),
+    # 'ballot styles' is not a list
+    (
+        {
+            "name": "General Election",
+            "type": "general",
+            "start_date": "2024-11-05",
+            "end_date": "2024-11-05",
+            "ballot_styles": {},
+        },
+        raises(TypeError, match = "Field 'ballot_styles' is not of type 'List'"),
+    ),
+]
+
+
+@pytest.mark.parametrize("data, raises", ELECTION_TESTS)
+def test_election(data, raises):
+    with raises:
+        item = ElectionData(**data)
+
+
+def test_election_fields():
+    data = {
+        "name": "General Election",
+        "type": "general",
+        "start_date": "2024-11-05",
+        "end_date": "2024-11-05",
+        "ballot_styles": [
+            {
+                "id": "precinct_2_spacetown",
+                "scopes": [
+                    "spacetown-precinct",
+                ],
+                "contests": [
+                    {
+                        "id": "candidate-contest-orbit-city-mayor",
+                        "title": "Mayor of Orbit City",
+                        "type": "candidate",
+                        "vote_type": "plurality",
+                        "votes_allowed": 1,
+                        "district": "Orbit City",
+                        "candidates": [
+                            {
+                                "id": "candidate-choice-1",
+                                "name": [
+                                    "Spencer Cogswell"
+                                ],
+                                "party": [
+                                    {
+                                        "name": "The Hadron Party of Farallon",
+                                        "abbreviation": "HAD"
+                                    },
+                                ],
+                                "is_write_in": False,
+                            },
+                        ],
+                    },
+                ],
+            },
+        ],
+    }
+    item = ElectionData(**data)
+    # Scalar fields match
+    assert item.name == data["name"]
+    assert item.type == data["type"]
+    assert item.start_date == data["start_date"]
+    assert item.end_date == data["end_date"]
+    # Not the same type: data model converts each party to an object
+    assert item.ballot_styles != data["ballot_styles"]
+    # Lengths and fields are the same.
+    assert len(item.ballot_styles) == len(data["ballot_styles"])
+    assert all(isinstance(_, BallotStyleData) for _ in item.ballot_styles)
+    for actual, expected in zip(item.ballot_styles, data["ballot_styles"]):
+        actual = asdict(actual)
+        assert actual == expected

From 893af78666cc6da84d34f1375e4e76bc97a9c955 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Mon, 12 Sep 2022 10:35:08 -0700
Subject: [PATCH 38/48] Use a base type for contest models. Partly to share
 fields but primarily for correctness: contests have to be a single list
 independent of sub-type because that's how EDFs work. Splitting the types
 into multiple lists would lose the order.

---
 src/electos/ballotmaker/data/models.py | 59 +++++++++++++++-----------
 tests/models/test_ballot_style.py      |  5 ++-
 2 files changed, 37 insertions(+), 27 deletions(-)

diff --git a/src/electos/ballotmaker/data/models.py b/src/electos/ballotmaker/data/models.py
index 20c32f0..5d38cf4 100644
--- a/src/electos/ballotmaker/data/models.py
+++ b/src/electos/ballotmaker/data/models.py
@@ -1,6 +1,7 @@
 """Ballot data models."""
 
 from dataclasses import dataclass
+from enum import Enum
 from typing import List, Union
 
 
@@ -37,6 +38,31 @@ def _check_type_list(instance, field, type_):
         )
 
 
+# --- Contest model base type
+
+@dataclass
+class ContestData:
+
+    """Shared data for contests."""
+
+    id: str
+    type: str
+    title: str
+    district: str
+
+    def __post_init__(self):
+        _check_type(self, "id", str)
+        _check_type(self, "type", str)
+        _check_type(self, "title", str)
+        _check_type(self, "district", str)
+
+
+class ContestType(Enum):
+
+    BALLOT_MEASURE = "ballot measure"
+    CANDIDATE = "candidate"
+
+
 # --- Ballot contest data models
 
 
@@ -48,29 +74,21 @@ class BallotChoiceData:
     id: str
     choice: str
 
-
     def __post_init__(self):
         _check_type(self, "id", str)
         _check_type(self, "choice", str)
 
 
 @dataclass
-class BallotMeasureContestData:
+class BallotMeasureContestData(ContestData):
 
     """Data for ballot measure contests."""
 
-    id: str
-    type: str
-    title: str
-    district: str
     text: str
     choices: List[BallotChoiceData]
 
     def __post_init__(self):
-        _check_type(self, "id", str)
-        _check_type(self, "type", str)
-        _check_type(self, "title", str)
-        _check_type(self, "district", str)
+        super().__post_init__()
         _check_type(self, "text", str)
         _check_type_hint(self, "choices", List)
         self.choices = [BallotChoiceData(**_) for _ in self.choices]
@@ -112,23 +130,16 @@ def __post_init__(self):
 
 
 @dataclass
-class CandidateContestData:
+class CandidateContestData(ContestData):
 
     """Data for candidate contests."""
 
-    id: str
-    type: str
-    title: str
-    district: str
     vote_type: str
     votes_allowed: str
     candidates: List[CandidateChoiceData]
 
     def __post_init__(self):
-        _check_type(self, "id", str)
-        _check_type(self, "type", str)
-        _check_type(self, "title", str)
-        _check_type(self, "district", str)
+        super().__post_init__()
         _check_type(self, "vote_type", str)
         _check_type(self, "votes_allowed", int)
         _check_type_hint(self, "candidates", List)
@@ -140,9 +151,6 @@ class BallotStyleData:
 
     """Date for contests associated with a ballot style."""
 
-    BALLOT_MEASURE = "ballot measure"
-    CANDIDATE = "candidate"
-
     # Note: Don't use separate fields for the types of contests.
     # There's no guarantee the types will be clearly separated in an EDF.
     # (The NIST SP-1500-100 JSON Schema uses unions too.)
@@ -157,16 +165,17 @@ def __post_init__(self):
         _check_type_list(self, "scopes", str)
         _check_type_hint(self, "contests", List)
         contests = []
+        contest_types = [_.value for _ in ContestType]
         for contest in self.contests:
             if not isinstance(contest, dict):
                 raise TypeError(f"Contest is not a dictionary: '{contest}'")
             if "type" not in contest:
                 raise KeyError(f"Contest has no 'type' field: '{contest}'")
-            if contest["type"] not in (self.BALLOT_MEASURE, self.CANDIDATE):
+            if contest["type"] not in contest_types:
                 raise ValueError(f"Unhandled contest type: '{contest['type']}'")
-            if contest["type"] == self.BALLOT_MEASURE:
+            if contest["type"] == ContestType.BALLOT_MEASURE.value:
                 contest = BallotMeasureContestData(**contest)
-            elif contest["type"] == self.CANDIDATE:
+            elif contest["type"] == ContestType.CANDIDATE.value:
                 contest = CandidateContestData(**contest)
             contests.append(contest)
         self.contests = contests
diff --git a/tests/models/test_ballot_style.py b/tests/models/test_ballot_style.py
index eac2e0b..82ef6d8 100644
--- a/tests/models/test_ballot_style.py
+++ b/tests/models/test_ballot_style.py
@@ -9,6 +9,7 @@
     BallotMeasureContestData,
     BallotStyleData,
     CandidateContestData,
+    ContestType,
 )
 
 
@@ -403,7 +404,7 @@
             "scopes": [
                 "spacetown-precinct",
             ],
-            "contests": [{ "type": BallotStyleData.BALLOT_MEASURE }],
+            "contests": [{ "type": ContestType.BALLOT_MEASURE.value }],
         },
         raises(TypeError, match = "missing 5 required positional arguments: 'id', 'title', 'district', 'text', and 'choices'"),
     ),
@@ -414,7 +415,7 @@
             "scopes": [
                 "spacetown-precinct",
             ],
-            "contests": [{ "type": BallotStyleData.CANDIDATE }],
+            "contests": [{ "type": ContestType.CANDIDATE.value }],
         },
         raises(TypeError, match = "missing 6 required positional arguments: 'id', 'title', 'district', 'vote_type', 'votes_allowed', and 'candidates'"),
     ),

From e03c3bbd78b34ae4fa0604228ef04ccd66615e73 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Mon, 12 Sep 2022 13:34:54 -0700
Subject: [PATCH 39/48] Add derived candidate and ballot measure contest
 properties for ballot styles.

---
 src/electos/ballotmaker/data/models.py | 15 +++++++++++++++
 tests/models/test_ballot_style.py      |  2 ++
 2 files changed, 17 insertions(+)

diff --git a/src/electos/ballotmaker/data/models.py b/src/electos/ballotmaker/data/models.py
index 5d38cf4..272dd00 100644
--- a/src/electos/ballotmaker/data/models.py
+++ b/src/electos/ballotmaker/data/models.py
@@ -2,6 +2,7 @@
 
 from dataclasses import dataclass
 from enum import Enum
+from functools import cached_property
 from typing import List, Union
 
 
@@ -181,6 +182,20 @@ def __post_init__(self):
         self.contests = contests
 
 
+    @cached_property
+    def ballot_measure_contests(self):
+        return [
+            _ for _ in self.contests if _.type == ContestType.BALLOT_MEASURE.value
+        ]
+
+
+    @cached_property
+    def candidate_contests(self):
+        return [
+            _ for _ in self.contests if _.type == ContestType.CANDIDATE.value
+        ]
+
+
 @dataclass
 class ElectionData:
 
diff --git a/tests/models/test_ballot_style.py b/tests/models/test_ballot_style.py
index 82ef6d8..4de635b 100644
--- a/tests/models/test_ballot_style.py
+++ b/tests/models/test_ballot_style.py
@@ -487,6 +487,8 @@ def test_ballot_style_fields():
     # Types are mixed
     assert isinstance(item.contests[0], CandidateContestData)
     assert isinstance(item.contests[1], BallotMeasureContestData)
+    assert item.candidate_contests == [item.contests[0]]
+    assert item.ballot_measure_contests == [item.contests[1]]
     for actual, expected in zip(item.contests, data["contests"]):
         actual = asdict(actual)
         assert actual == expected

From a7f19c1b6d7b572364fbbfaa3997284083b03b8d Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Mon, 12 Sep 2022 16:00:58 -0700
Subject: [PATCH 40/48] Reorder contest fields in ballot data extractor to
 match ballot data model. Do some minor code cleanup.

---
 src/electos/ballotmaker/data/extractor.py | 11 ++++++-----
 1 file changed, 6 insertions(+), 5 deletions(-)

diff --git a/src/electos/ballotmaker/data/extractor.py b/src/electos/ballotmaker/data/extractor.py
index c0bf0b9..9813063 100644
--- a/src/electos/ballotmaker/data/extractor.py
+++ b/src/electos/ballotmaker/data/extractor.py
@@ -70,6 +70,7 @@ def ballot_style_external_id(ballot_style: BallotStyle):
 
 
 def ballot_style_gp_units(ballot_style: BallotStyle, index):
+    """Yield geo-political units for a ballot style."""
     for id_ in ballot_style.gp_unit_ids:
         gp_unit = index.by_id(id_)
         yield gp_unit
@@ -91,7 +92,7 @@ def candidate_name(candidate: Candidate):
 def candidate_party(candidate: Candidate, index):
     """Get the name and abbreviation of the party of a candidate as it appears on a ballot.
 
-       Drop either field from result if it isn't present.
+    Drop either field from result if it isn't present.
     """
     # Note: party ID is returned to allow de-duplicating parties in callers.
     id_ = candidate.party_id
@@ -193,12 +194,12 @@ def extract_candidate_contest(contest: CandidateContest, index):
     candidates = candidate_contest_candidates(contest, index)
     result = {
         "id": contest.model__id,
-        "title": contest.name,
         "type": "candidate",
+        "title": contest.name,
+        "district": district,
         "vote_type": contest.vote_variation.value,
         # Include even when default is 1: don't require caller to track that.
         "votes_allowed": contest.votes_allowed,
-        "district": district,
         "candidates": candidates,
         # "offices": offices,
         # "parties": parties,
@@ -221,11 +222,11 @@ def extract_ballot_measure_contest(contest: BallotMeasureContest, index):
     full_text = text_content(contest.full_text)
     result = {
         "id": contest.model__id,
-        "title": contest.name,
         "type": "ballot measure",
+        "title": contest.name,
         "district": district,
-        "choices": choices,
         "text": full_text,
+        "choices": choices,
     }
     return result
 

From 341df82d9f443475268b453f3dfe15fff43c1a64 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Mon, 12 Sep 2022 16:18:38 -0700
Subject: [PATCH 41/48] Don't return an empty party dictionary when there is no
 party. An empty dictionary adds a special case to 'PartyData' construction.

---
 src/electos/ballotmaker/data/extractor.py | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/src/electos/ballotmaker/data/extractor.py b/src/electos/ballotmaker/data/extractor.py
index 9813063..b3d7ab4 100644
--- a/src/electos/ballotmaker/data/extractor.py
+++ b/src/electos/ballotmaker/data/extractor.py
@@ -137,8 +137,9 @@ def candidate_contest_candidates(contest: CandidateContest, index):
                 if name:
                     names.append(name)
                 party, _party_id = candidate_party(candidate, index)
-                parties.append(party)
-                _party_ids.add(_party_id)
+                if party:
+                    parties.append(party)
+                    _party_ids.add(_party_id)
         # If there's only one party ID, all candidates share the same party.
         # If there's any divergence track them all individually.
         if len(_party_ids) == 1:

From 002e41564449972aac61cfef97638a50c41b135d Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Mon, 12 Sep 2022 18:01:13 -0700
Subject: [PATCH 42/48] Re-combine candidate and ballot measure contests in
 extractor.

---
 src/electos/ballotmaker/data/edf-to-ballot-data.py | 2 +-
 src/electos/ballotmaker/data/extractor.py          | 7 +------
 2 files changed, 2 insertions(+), 7 deletions(-)

diff --git a/src/electos/ballotmaker/data/edf-to-ballot-data.py b/src/electos/ballotmaker/data/edf-to-ballot-data.py
index a7d0634..06feb26 100644
--- a/src/electos/ballotmaker/data/edf-to-ballot-data.py
+++ b/src/electos/ballotmaker/data/edf-to-ballot-data.py
@@ -22,7 +22,7 @@ def report(root, index, nth, **opts):
     data = {}
     id_ = ballot_style_external_id(ballot_style)
     data["ballot_style"] = id_
-    contests = extract_contests(ballot_style, index)
+    contests = list(extract_contests(ballot_style, index))
     if not contests:
         print(f"No contests found for ballot style: {id_}\n")
     data["contests"] = contests
diff --git a/src/electos/ballotmaker/data/extractor.py b/src/electos/ballotmaker/data/extractor.py
index b3d7ab4..ad025a3 100644
--- a/src/electos/ballotmaker/data/extractor.py
+++ b/src/electos/ballotmaker/data/extractor.py
@@ -234,20 +234,15 @@ def extract_ballot_measure_contest(contest: BallotMeasureContest, index):
 
 def extract_contests(ballot_style: BallotStyle, index):
     """Extract contest subset needed for ballots."""
-    contests = {
-        kind: [] for kind in ("candidate", "ballot_measure")
-    }
     for contest in ballot_style_contests(ballot_style, index):
         if isinstance(contest, CandidateContest):
             entry = extract_candidate_contest(contest, index)
-            contests["candidate"].append(entry)
         elif isinstance(contest, BallotMeasureContest):
             entry = extract_ballot_measure_contest(contest, index)
-            contests["ballot_measure"].append(entry)
         else:
             # Ignore other contest types
             print(f"Skipping contest of type {contest.model__type}")
-    return contests
+        yield entry
 
 
 def extract_ballot_styles(election_report: ElectionReport, index):

From b173842306590f842fcde1b88e33875ec8f54ce3 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Tue, 13 Sep 2022 19:42:20 -0700
Subject: [PATCH 43/48] Re-implement and simplify the ballot extractor
 interface. - Unify extraction functions:   - Use direct field accesses for
 properties, index for references.   - Yield data dictionaries instead of
 returning lists. - Add in election data. - Entry point is now
 'extract_ballot_data', which returns a list.

---
 .../ballotmaker/data/edf-to-ballot-data.py    | 31 ++++------------
 src/electos/ballotmaker/data/extractor.py     | 36 +++++++++++++++++--
 2 files changed, 39 insertions(+), 28 deletions(-)

diff --git a/src/electos/ballotmaker/data/edf-to-ballot-data.py b/src/electos/ballotmaker/data/edf-to-ballot-data.py
index 06feb26..56327c0 100644
--- a/src/electos/ballotmaker/data/edf-to-ballot-data.py
+++ b/src/electos/ballotmaker/data/edf-to-ballot-data.py
@@ -5,27 +5,12 @@
 from electos.datamodels.nist.indexes import ElementIndex
 from electos.datamodels.nist.models.edf import ElectionReport
 
-from electos.ballotmaker.data.extractor import (
-    ballot_style_external_id,
-    extract_ballot_styles,
-    extract_contests,
-)
+from electos.ballotmaker.data.extractor import extract_ballot_data
 
 
-def report(root, index, nth, **opts):
+def report(document, index, **opts):
     """Generate data needed by BallotLab."""
-    ballot_styles = list(extract_ballot_styles(root, index))
-    if not (1 <= nth <= len(ballot_styles)):
-        print(f"Ballot styles: {nth} is out of range [1-{len(ballot_styles)}]")
-        return
-    ballot_style = ballot_styles[nth - 1]
-    data = {}
-    id_ = ballot_style_external_id(ballot_style)
-    data["ballot_style"] = id_
-    contests = list(extract_contests(ballot_style, index))
-    if not contests:
-        print(f"No contests found for ballot style: {id_}\n")
-    data["contests"] = contests
+    data = extract_ballot_data(document, index)
     print(json.dumps(data, indent = 4))
 
 
@@ -35,10 +20,6 @@ def main():
         "file", type = Path,
         help = "Test case data (JSON)"
     )
-    parser.add_argument(
-        "nth", nargs = "?", type = int, default = 1,
-        help = "Index of the ballot style, starting from 1 (default: 1)"
-    )
     parser.add_argument(
         "--debug", action = "store_true",
         help = "Enable debugging output and stack traces"
@@ -51,9 +32,9 @@ def main():
         with file.open() as input:
             text = input.read()
             data = json.loads(text)
-        edf = ElectionReport(**data)
-        index = ElementIndex(edf, "ElectionResults")
-        report(edf, index, **opts)
+        document = ElectionReport(**data)
+        index = ElementIndex(document, "ElectionResults")
+        report(document, index, **opts)
     except Exception as ex:
         if opts["debug"]:
             raise ex
diff --git a/src/electos/ballotmaker/data/extractor.py b/src/electos/ballotmaker/data/extractor.py
index ad025a3..21a3387 100644
--- a/src/electos/ballotmaker/data/extractor.py
+++ b/src/electos/ballotmaker/data/extractor.py
@@ -8,6 +8,7 @@
     Candidate,
     CandidateContest,
     CandidateSelection,
+    Election,
     ElectionReport,
     InternationalizedText,
     OrderedContest,
@@ -245,7 +246,36 @@ def extract_contests(ballot_style: BallotStyle, index):
         yield entry
 
 
-def extract_ballot_styles(election_report: ElectionReport, index):
+def extract_ballot_styles(election: Election, index):
     """Extract all ballot styles."""
-    for ballot_style in index.by_type("BallotStyle"):
-        yield ballot_style
+    for ballot_style in election.ballot_style:
+        data = {
+            "id": ballot_style_external_id(ballot_style),
+            "scopes": [ _.model__id for _ in ballot_style_gp_units(ballot_style, index) ],
+            "contests": [ _ for _ in extract_contests(ballot_style, index) ],
+        }
+        yield data
+
+
+def extract_election_data(election_report: ElectionReport, index):
+    """Extract all elections."""
+    # In most cases there isn't more than one 'Election' in a report, but the
+    # standard allows more than one, so handle them.
+    for election in election_report.election:
+        data = {
+            "name": text_content(election.name),
+            "type": election.type.value,
+            "start_date": election.start_date.strftime("%Y-%m-%d"),
+            "end_date": election.end_date.strftime("%Y-%m-%d"),
+            "ballot_styles": [_ for _ in extract_ballot_styles(election, index)],
+        }
+        yield data
+
+
+def extract_ballot_data(election_report: ElectionReport, index: ElementIndex):
+    """Extract election data.
+    This is the primary entry point for the extractor.
+    """
+    index = index or ElementIndex(election_report, "ElectionResults")
+    election_data = [ _ for _ in extract_election_data(election_report, index) ]
+    return election_data

From 5ace8d6123143f666a514b158d55ef3328d616c8 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Tue, 13 Sep 2022 20:04:54 -0700
Subject: [PATCH 44/48] Move document and index creation into 'report'.

---
 src/electos/ballotmaker/data/edf-to-ballot-data.py | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/src/electos/ballotmaker/data/edf-to-ballot-data.py b/src/electos/ballotmaker/data/edf-to-ballot-data.py
index 56327c0..0d68c5c 100644
--- a/src/electos/ballotmaker/data/edf-to-ballot-data.py
+++ b/src/electos/ballotmaker/data/edf-to-ballot-data.py
@@ -8,10 +8,12 @@
 from electos.ballotmaker.data.extractor import extract_ballot_data
 
 
-def report(document, index, **opts):
+def report(data, **opts):
     """Generate data needed by BallotLab."""
-    data = extract_ballot_data(document, index)
-    print(json.dumps(data, indent = 4))
+    document = ElectionReport(**data)
+    index = ElementIndex(document, "ElectionResults")
+    ballot_data = extract_ballot_data(document, index)
+    print(json.dumps(ballot_data, indent = 4))
 
 
 def main():
@@ -32,9 +34,7 @@ def main():
         with file.open() as input:
             text = input.read()
             data = json.loads(text)
-        document = ElectionReport(**data)
-        index = ElementIndex(document, "ElectionResults")
-        report(document, index, **opts)
+        report(data, **opts)
     except Exception as ex:
         if opts["debug"]:
             raise ex

From 04e0d3e1bf11dc2aefb7509f01895c30ab51773a Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Tue, 13 Sep 2022 20:16:45 -0700
Subject: [PATCH 45/48] Make 'ElectionReport' and 'ElementIndex' internal to
 the extractor. The caller only receives an 'ElectionData' back.

---
 src/electos/ballotmaker/data/edf-to-ballot-data.py |  9 +++------
 src/electos/ballotmaker/data/extractor.py          | 10 +++++++---
 2 files changed, 10 insertions(+), 9 deletions(-)

diff --git a/src/electos/ballotmaker/data/edf-to-ballot-data.py b/src/electos/ballotmaker/data/edf-to-ballot-data.py
index 0d68c5c..0647722 100644
--- a/src/electos/ballotmaker/data/edf-to-ballot-data.py
+++ b/src/electos/ballotmaker/data/edf-to-ballot-data.py
@@ -1,18 +1,15 @@
 import argparse
 import json
+from dataclasses import asdict
 from pathlib import Path
 
-from electos.datamodels.nist.indexes import ElementIndex
-from electos.datamodels.nist.models.edf import ElectionReport
-
 from electos.ballotmaker.data.extractor import extract_ballot_data
 
 
 def report(data, **opts):
     """Generate data needed by BallotLab."""
-    document = ElectionReport(**data)
-    index = ElementIndex(document, "ElectionResults")
-    ballot_data = extract_ballot_data(document, index)
+    ballot_data = extract_ballot_data(data)
+    ballot_data = [asdict(_) for _ in ballot_data]
     print(json.dumps(ballot_data, indent = 4))
 
 
diff --git a/src/electos/ballotmaker/data/extractor.py b/src/electos/ballotmaker/data/extractor.py
index 21a3387..e157cc5 100644
--- a/src/electos/ballotmaker/data/extractor.py
+++ b/src/electos/ballotmaker/data/extractor.py
@@ -1,4 +1,4 @@
-from typing import List, Union
+from typing import Dict, List, Union
 
 from electos.datamodels.nist.indexes import ElementIndex
 from electos.datamodels.nist.models.edf import (
@@ -272,10 +272,14 @@ def extract_election_data(election_report: ElectionReport, index):
         yield data
 
 
-def extract_ballot_data(election_report: ElectionReport, index: ElementIndex):
+def extract_ballot_data(data: Dict, index: ElementIndex = None) -> ElectionData:
     """Extract election data.
+
     This is the primary entry point for the extractor.
     """
+    election_report = ElectionReport(**data)
     index = index or ElementIndex(election_report, "ElectionResults")
-    election_data = [ _ for _ in extract_election_data(election_report, index) ]
+    election_data = [
+        ElectionData(**_) for _ in extract_election_data(election_report, index)
+    ]
     return election_data

From 5350723a2871f53f07c14a193f55f14749c08886 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Tue, 13 Sep 2022 20:18:09 -0700
Subject: [PATCH 46/48] Document the entry point of the extractor API.

---
 src/electos/ballotmaker/data/extractor.py | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/src/electos/ballotmaker/data/extractor.py b/src/electos/ballotmaker/data/extractor.py
index e157cc5..8001a36 100644
--- a/src/electos/ballotmaker/data/extractor.py
+++ b/src/electos/ballotmaker/data/extractor.py
@@ -276,6 +276,15 @@ def extract_ballot_data(data: Dict, index: ElementIndex = None) -> ElectionData:
     """Extract election data.
 
     This is the primary entry point for the extractor.
+
+    Parameters:
+        data: An EDF / election report dictionary.
+        index: An ElementIndex.
+            If empty (the default), create a new index from the election report.
+            Use this parameter only if there's already an existing index.
+
+    Returns:
+        Election data model for use in ballot rendering.
     """
     election_report = ElectionReport(**data)
     index = index or ElementIndex(election_report, "ElectionResults")

From 7e2f748d74d9a329dbb615776c74c8623aa84673 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Tue, 13 Sep 2022 20:40:18 -0700
Subject: [PATCH 47/48] Turn ballot extractor into a class. - Public API is now
 'extract'. - Prefix all internal methods as private.   - Prefix utility
 functions but don't make them methods. - Make the element index an instance
 variable. - Re-wrap text to fit within line length.

Note: In a class method order usually puts public functions at the top.
Not doing that here to avoid making the change diff useless.
---
 .../ballotmaker/data/edf-to-ballot-data.py    |   5 +-
 src/electos/ballotmaker/data/extractor.py     | 479 +++++++++---------
 2 files changed, 251 insertions(+), 233 deletions(-)

diff --git a/src/electos/ballotmaker/data/edf-to-ballot-data.py b/src/electos/ballotmaker/data/edf-to-ballot-data.py
index 0647722..42963a7 100644
--- a/src/electos/ballotmaker/data/edf-to-ballot-data.py
+++ b/src/electos/ballotmaker/data/edf-to-ballot-data.py
@@ -3,12 +3,13 @@
 from dataclasses import asdict
 from pathlib import Path
 
-from electos.ballotmaker.data.extractor import extract_ballot_data
+from electos.ballotmaker.data.extractor import BallotDataExtractor
 
 
 def report(data, **opts):
     """Generate data needed by BallotLab."""
-    ballot_data = extract_ballot_data(data)
+    extractor = BallotDataExtractor()
+    ballot_data = extractor.extract(data)
     ballot_data = [asdict(_) for _ in ballot_data]
     print(json.dumps(ballot_data, indent = 4))
 
diff --git a/src/electos/ballotmaker/data/extractor.py b/src/electos/ballotmaker/data/extractor.py
index 8001a36..10463e0 100644
--- a/src/electos/ballotmaker/data/extractor.py
+++ b/src/electos/ballotmaker/data/extractor.py
@@ -1,5 +1,6 @@
 from typing import Dict, List, Union
 
+from electos.ballotmaker.data.models import ElectionData
 from electos.datamodels.nist.indexes import ElementIndex
 from electos.datamodels.nist.models.edf import (
     BallotMeasureContest,
@@ -28,267 +29,283 @@
 # --- Utilities
 
 
-def text_content(item):
+def _text_content(item):
     """Return joined lines from internationalized text."""
     assert isinstance(item, InternationalizedText)
     text = "\n".join(_.content for _ in item.text)
     return text
 
 
-def walk_ordered_contests(content: List[OrderedContent]):
+def _walk_ordered_contests(content: List[OrderedContent]):
     """Walk ordered content yielding contests."""
     for item in content:
         if isinstance(item, OrderedContest):
             yield item
         elif isinstance(item, OrderedHeader):
-            yield from walk_ordered_contests(item.ordered_content)
+            yield from _walk_ordered_contests(item.ordered_content)
         else:
             raise TypeError(f"Unexpected type: {type(item).__name__}")
 
 
-def walk_ordered_headers(content: List[OrderedContent]):
+def _walk_ordered_headers(content: List[OrderedContent]):
     """Walk ordered content yielding headers."""
     for item in content:
         if isinstance(item, OrderedHeader):
             yield item
-            yield from walk_ordered_headers(item.ordered_content)
+            yield from _walk_ordered_headers(item.ordered_content)
         else:
             raise TypeError(f"Unexpected type: {type(item).__name__}")
 
 
-# --- Ballot Properties
+# --- Extractor
 
+class BallotDataExtractor:
 
-def ballot_style_external_id(ballot_style: BallotStyle):
-    """Get the text of a ballot style's external identifier if any."""
-    if ballot_style.external_identifier:
-        assert len(ballot_style.external_identifier) == 1, \
-            "Not ready to handle multiple BallotStyle external IDs"
-        name = ballot_style.external_identifier[0].value
-    else:
-        name = ""
-    return name
+    """Extract election data from an EDF."""
 
 
-def ballot_style_gp_units(ballot_style: BallotStyle, index):
-    """Yield geo-political units for a ballot style."""
-    for id_ in ballot_style.gp_unit_ids:
-        gp_unit = index.by_id(id_)
-        yield gp_unit
+    def __init__(self):
+        pass
 
 
-def ballot_style_contests(ballot_style: BallotStyle, index):
-    """Yield the contests of a ballot style."""
-    for item in walk_ordered_contests(ballot_style.ordered_content):
-        contest = index.by_id(item.contest_id)
-        yield contest
-
-
-def candidate_name(candidate: Candidate):
-    """Get the name of a candidate as it appears on a ballot."""
-    name = text_content(candidate.ballot_name)
-    return name
-
-
-def candidate_party(candidate: Candidate, index):
-    """Get the name and abbreviation of the party of a candidate as it appears on a ballot.
-
-    Drop either field from result if it isn't present.
-    """
-    # Note: party ID is returned to allow de-duplicating parties in callers.
-    id_ = candidate.party_id
-    party = index.by_id(id_)
-    name = text_content(party.name) if party else None
-    abbreviation = text_content(party.abbreviation) if party and party.abbreviation else None
-    result = {}
-    if name:
-        result["name"] = name
-    if abbreviation:
-        result["abbreviation"] = abbreviation
-    return result, id_
-
-
-def candidate_contest_candidates(contest: CandidateContest, index):
-    """Get candidates for contest, grouped by slate/ticket.
-
-    A slate has:
-
-    - A single ID for the contest selection
-    - Collects candidate names into an array.
-    - Collects candidate parties into an array.
-      - If all candidates in a race share a single party they are combined into
-        one entry in the array.
-      - If any candidates differ from the others, parties are listed separately.
-
-    Notes:
-        - There's no clear guarantee of a 1:1 relationship between slates and parties.
-    """
-    # Collect individual candidates
-    candidates = []
-    for selection in contest.contest_selection:
-        assert isinstance(selection, CandidateSelection), \
-            f"Unexpected non-candidate selection: {type(selection).__name__}"
-        names = []
+    def _ballot_style_external_id(self, ballot_style: BallotStyle):
+        """Get the text of a ballot style's external identifier if any."""
+        if ballot_style.external_identifier:
+            assert len(ballot_style.external_identifier) == 1, \
+                "Not ready to handle multiple BallotStyle external IDs"
+            name = ballot_style.external_identifier[0].value
+        else:
+            name = ""
+        return name
+
+
+    def _ballot_style_gp_units(self, ballot_style: BallotStyle):
+        """Yield geo-political units for a ballot style."""
+        for id_ in ballot_style.gp_unit_ids:
+            gp_unit = self._index.by_id(id_)
+            yield gp_unit
+
+
+    def _ballot_style_contests(self, ballot_style: BallotStyle):
+        """Yield the contests of a ballot style."""
+        for item in _walk_ordered_contests(ballot_style.ordered_content):
+            contest = self._index.by_id(item.contest_id)
+            yield contest
+
+
+    def _candidate_name(self, candidate: Candidate):
+        """Get the name of a candidate as it appears on a ballot."""
+        name = _text_content(candidate.ballot_name)
+        return name
+
+
+    def _candidate_party(self, candidate: Candidate):
+        """Get the name and abbreviation of the party of a candidate as it appears on a ballot.
+
+        Drop either field from result if it isn't present.
+        """
+        # Note: party ID is returned to allow de-duplicating parties in callers.
+        id_ = candidate.party_id
+        party = self._index.by_id(id_)
+        name = _text_content(party.name) if party else None
+        abbreviation = (
+            _text_content(party.abbreviation) if party and party.abbreviation else None
+        )
+        result = {}
+        if name:
+            result["name"] = name
+        if abbreviation:
+            result["abbreviation"] = abbreviation
+        return result, id_
+
+
+    def _candidate_contest_candidates(self, contest: CandidateContest):
+        """Get candidates for contest, grouped by slate/ticket.
+
+        A slate has:
+
+        - A single ID for the contest selection
+        - Collects candidate names into an array.
+        - Collects candidate parties into an array.
+          - If all candidates in a race share a single party they are combined into
+            one entry in the array.
+          - If any candidates differ from the others, parties are listed separately.
+
+        Notes:
+            - There's no clear guarantee of a 1:1 relationship between slates/tickets
+              and parties.
+        """
+        # Collect individual candidates
+        candidates = []
+        for selection in contest.contest_selection:
+            assert isinstance(selection, CandidateSelection), \
+                f"Unexpected non-candidate selection: {type(selection).__name__}"
+            names = []
+            parties = []
+            _party_ids = set()
+            if selection.candidate_ids:
+                for id_ in selection.candidate_ids:
+                    candidate = self._index.by_id(id_)
+                    name = self._candidate_name(candidate)
+                    if name:
+                        names.append(name)
+                    party, _party_id = self._candidate_party(candidate)
+                    if party:
+                        parties.append(party)
+                        _party_ids.add(_party_id)
+            # If there's only one party ID, all candidates share the same party.
+            # If there's any divergence track them all individually.
+            if len(_party_ids) == 1:
+                parties = parties[:1]
+            result = {
+                "id": selection.model__id,
+                "name": names,
+                "party": parties,
+                "is_write_in": bool(selection.is_write_in)
+            }
+            candidates.append(result)
+        return candidates
+
+
+    def _candidate_contest_offices(self, contest: CandidateContest):
+        """Get any offices associated with a candidate contest."""
+        offices = []
+        if contest.office_ids:
+            for id_ in contest.office_ids:
+                office = self._index.by_id(id_)
+                name = _text_content(office.name)
+                offices.append(name)
+        return offices
+
+
+    def _candidate_contest_parties(self, contest: CandidateContest):
+        """Get any parties associated with a candidate contest."""
         parties = []
-        _party_ids = set()
-        if selection.candidate_ids:
-            for id_ in selection.candidate_ids:
-                candidate = index.by_id(id_)
-                name = candidate_name(candidate)
-                if name:
-                    names.append(name)
-                party, _party_id = candidate_party(candidate, index)
-                if party:
-                    parties.append(party)
-                    _party_ids.add(_party_id)
-        # If there's only one party ID, all candidates share the same party.
-        # If there's any divergence track them all individually.
-        if len(_party_ids) == 1:
-            parties = parties[:1]
+        if contest.primary_party_ids:
+            for id_ in contest.primary_party_ids:
+                party = self._index.by_id(id_)
+                name = _text_content(party.name)
+                parties.append(name)
+        return parties
+
+
+    def _contest_election_district(self, contest: Contest):
+        """Get the district name of a contest."""
+        district = self._index.by_id(contest.election_district_id)
+        district = _text_content(district.name)
+        return district
+
+
+    def _extract_candidate_contest(self, contest: CandidateContest):
+        """Extract candidate contest subset needed for a ballot."""
+        district = self._contest_election_district(contest)
+        offices = self._candidate_contest_offices(contest)
+        parties = self._candidate_contest_parties(contest)
+        candidates = self._candidate_contest_candidates(contest)
         result = {
-            "id": selection.model__id,
-            "name": names,
-            "party": parties,
-            "is_write_in": bool(selection.is_write_in)
-        }
-        candidates.append(result)
-    return candidates
-
-
-def candidate_contest_offices(contest: CandidateContest, index):
-    """Get any offices associated with a candidate contest."""
-    offices = []
-    if contest.office_ids:
-        for id_ in contest.office_ids:
-            office = index.by_id(id_)
-            name = text_content(office.name)
-            offices.append(name)
-    return offices
-
-
-def candidate_contest_parties(contest: CandidateContest, index):
-    """Get any parties associated with a candidate contest."""
-    parties = []
-    if contest.primary_party_ids:
-        for id_ in contest.primary_party_ids:
-            party = index.by_id(id_)
-            name = text_content(party.name)
-            parties.append(name)
-    return parties
-
-
-def contest_election_district(contest: Contest, index):
-    """Get the district name of a contest."""
-    district = index.by_id(contest.election_district_id)
-    district = text_content(district.name)
-    return district
-
-
-# --- Extraction
-#
-# Gather and select data needed for ballot generation.
-
-def extract_candidate_contest(contest: CandidateContest, index):
-    """Extract candidate contest subset needed for a ballot."""
-    district = contest_election_district(contest, index)
-    offices = candidate_contest_offices(contest, index)
-    parties = candidate_contest_parties(contest, index)
-    candidates = candidate_contest_candidates(contest, index)
-    result = {
-        "id": contest.model__id,
-        "type": "candidate",
-        "title": contest.name,
-        "district": district,
-        "vote_type": contest.vote_variation.value,
-        # Include even when default is 1: don't require caller to track that.
-        "votes_allowed": contest.votes_allowed,
-        "candidates": candidates,
-        # "offices": offices,
-        # "parties": parties,
-    }
-    return result
-
-
-def extract_ballot_measure_contest(contest: BallotMeasureContest, index):
-    """Extract ballot measure contest subset needed for a ballot."""
-    choices = []
-    for selection in contest.contest_selection:
-        assert isinstance(selection, BallotMeasureSelection), \
-           f"Unexpected non-ballot measure selection: {type(selection).__name__}"
-        choice = text_content(selection.selection)
-        choices.append({
-            "id": selection.model__id,
-            "choice": choice,
-        })
-    district = contest_election_district(contest, index)
-    full_text = text_content(contest.full_text)
-    result = {
-        "id": contest.model__id,
-        "type": "ballot measure",
-        "title": contest.name,
-        "district": district,
-        "text": full_text,
-        "choices": choices,
-    }
-    return result
-
-
-def extract_contests(ballot_style: BallotStyle, index):
-    """Extract contest subset needed for ballots."""
-    for contest in ballot_style_contests(ballot_style, index):
-        if isinstance(contest, CandidateContest):
-            entry = extract_candidate_contest(contest, index)
-        elif isinstance(contest, BallotMeasureContest):
-            entry = extract_ballot_measure_contest(contest, index)
-        else:
-            # Ignore other contest types
-            print(f"Skipping contest of type {contest.model__type}")
-        yield entry
-
-
-def extract_ballot_styles(election: Election, index):
-    """Extract all ballot styles."""
-    for ballot_style in election.ballot_style:
-        data = {
-            "id": ballot_style_external_id(ballot_style),
-            "scopes": [ _.model__id for _ in ballot_style_gp_units(ballot_style, index) ],
-            "contests": [ _ for _ in extract_contests(ballot_style, index) ],
+            "id": contest.model__id,
+            "type": "candidate",
+            "title": contest.name,
+            "district": district,
+            "vote_type": contest.vote_variation.value,
+            # Include even when default is 1: don't require caller to track that.
+            "votes_allowed": contest.votes_allowed,
+            "candidates": candidates,
+            # "offices": offices,
+            # "parties": parties,
         }
-        yield data
-
-
-def extract_election_data(election_report: ElectionReport, index):
-    """Extract all elections."""
-    # In most cases there isn't more than one 'Election' in a report, but the
-    # standard allows more than one, so handle them.
-    for election in election_report.election:
-        data = {
-            "name": text_content(election.name),
-            "type": election.type.value,
-            "start_date": election.start_date.strftime("%Y-%m-%d"),
-            "end_date": election.end_date.strftime("%Y-%m-%d"),
-            "ballot_styles": [_ for _ in extract_ballot_styles(election, index)],
+        return result
+
+
+    def _extract_ballot_measure_contest(self, contest: BallotMeasureContest):
+        """Extract ballot measure contest subset needed for a ballot."""
+        choices = []
+        for selection in contest.contest_selection:
+            assert isinstance(selection, BallotMeasureSelection), \
+               f"Unexpected non-ballot measure selection: {type(selection).__name__}"
+            choice = _text_content(selection.selection)
+            choices.append({
+                "id": selection.model__id,
+                "choice": choice,
+            })
+        district = self._contest_election_district(contest)
+        full_text = _text_content(contest.full_text)
+        result = {
+            "id": contest.model__id,
+            "type": "ballot measure",
+            "title": contest.name,
+            "district": district,
+            "text": full_text,
+            "choices": choices,
         }
-        yield data
-
-
-def extract_ballot_data(data: Dict, index: ElementIndex = None) -> ElectionData:
-    """Extract election data.
-
-    This is the primary entry point for the extractor.
-
-    Parameters:
-        data: An EDF / election report dictionary.
-        index: An ElementIndex.
-            If empty (the default), create a new index from the election report.
-            Use this parameter only if there's already an existing index.
-
-    Returns:
-        Election data model for use in ballot rendering.
-    """
-    election_report = ElectionReport(**data)
-    index = index or ElementIndex(election_report, "ElectionResults")
-    election_data = [
-        ElectionData(**_) for _ in extract_election_data(election_report, index)
-    ]
-    return election_data
+        return result
+
+
+    def _extract_contests(self, ballot_style: BallotStyle):
+        """Extract contest subset needed for ballots."""
+        for contest in self._ballot_style_contests(ballot_style):
+            if isinstance(contest, CandidateContest):
+                entry = self._extract_candidate_contest(contest)
+            elif isinstance(contest, BallotMeasureContest):
+                entry = self._extract_ballot_measure_contest(contest)
+            else:
+                # Ignore other contest types
+                print(f"Skipping contest of type {contest.model__type}")
+            yield entry
+
+
+    def _extract_ballot_styles(self, election: Election):
+        """Extract all ballot styles."""
+        for ballot_style in election.ballot_style:
+            data = {
+                "id": self._ballot_style_external_id(ballot_style),
+                "scopes": [
+                    _.model__id
+                    for _ in self._ballot_style_gp_units(ballot_style)
+                ],
+                "contests": [
+                    _
+                    for _ in self._extract_contests(ballot_style)
+                ],
+            }
+            yield data
+
+
+    def _extract_election_data(self, election_report: ElectionReport):
+        """Extract all elections."""
+        # In most cases there isn't more than one 'Election' in a report, but the
+        # standard allows more than one, so handle them.
+        for election in election_report.election:
+            data = {
+                "name": _text_content(election.name),
+                "type": election.type.value,
+                "start_date": election.start_date.strftime("%Y-%m-%d"),
+                "end_date": election.end_date.strftime("%Y-%m-%d"),
+                "ballot_styles": [
+                    _ for _ in self._extract_ballot_styles(election)
+                ],
+            }
+            yield data
+
+
+    def extract(self, data: Dict, index: ElementIndex = None) -> ElectionData:
+        """Extract election data.
+
+        This is the primary entry point for the extractor.
+
+        Parameters:
+            data: An EDF / election report dictionary.
+            index: An ElementIndex.
+                If empty (the default), create a new index from the election report.
+                Use this parameter only if there's already an existing index.
+
+        Returns:
+            Election data model for use in ballot rendering.
+        """
+        election_report = ElectionReport(**data)
+        self._index = index or ElementIndex(election_report, "ElectionResults")
+        election_data = [
+            ElectionData(**_)
+            for _ in self._extract_election_data(election_report)
+        ]
+        return election_data

From 1195709847099a378f3dc787187c6985575a5127 Mon Sep 17 00:00:00 2001
From: Ion Y <ion@osetinstitute.org>
Date: Tue, 13 Sep 2022 21:01:43 -0700
Subject: [PATCH 48/48] Rename extractor internal functions start with
 '_extract'.

---
 src/electos/ballotmaker/data/extractor.py | 20 ++++++++++----------
 1 file changed, 10 insertions(+), 10 deletions(-)

diff --git a/src/electos/ballotmaker/data/extractor.py b/src/electos/ballotmaker/data/extractor.py
index 10463e0..c4372da 100644
--- a/src/electos/ballotmaker/data/extractor.py
+++ b/src/electos/ballotmaker/data/extractor.py
@@ -196,7 +196,7 @@ def _contest_election_district(self, contest: Contest):
         return district
 
 
-    def _extract_candidate_contest(self, contest: CandidateContest):
+    def _candidate_contest_of(self, contest: CandidateContest):
         """Extract candidate contest subset needed for a ballot."""
         district = self._contest_election_district(contest)
         offices = self._candidate_contest_offices(contest)
@@ -217,7 +217,7 @@ def _extract_candidate_contest(self, contest: CandidateContest):
         return result
 
 
-    def _extract_ballot_measure_contest(self, contest: BallotMeasureContest):
+    def _ballot_measure_contest_of(self, contest: BallotMeasureContest):
         """Extract ballot measure contest subset needed for a ballot."""
         choices = []
         for selection in contest.contest_selection:
@@ -241,20 +241,20 @@ def _extract_ballot_measure_contest(self, contest: BallotMeasureContest):
         return result
 
 
-    def _extract_contests(self, ballot_style: BallotStyle):
+    def _contests(self, ballot_style: BallotStyle):
         """Extract contest subset needed for ballots."""
         for contest in self._ballot_style_contests(ballot_style):
             if isinstance(contest, CandidateContest):
-                entry = self._extract_candidate_contest(contest)
+                entry = self._candidate_contest_of(contest)
             elif isinstance(contest, BallotMeasureContest):
-                entry = self._extract_ballot_measure_contest(contest)
+                entry = self._ballot_measure_contest_of(contest)
             else:
                 # Ignore other contest types
                 print(f"Skipping contest of type {contest.model__type}")
             yield entry
 
 
-    def _extract_ballot_styles(self, election: Election):
+    def _election_ballot_styles(self, election: Election):
         """Extract all ballot styles."""
         for ballot_style in election.ballot_style:
             data = {
@@ -265,13 +265,13 @@ def _extract_ballot_styles(self, election: Election):
                 ],
                 "contests": [
                     _
-                    for _ in self._extract_contests(ballot_style)
+                    for _ in self._contests(ballot_style)
                 ],
             }
             yield data
 
 
-    def _extract_election_data(self, election_report: ElectionReport):
+    def _elections(self, election_report: ElectionReport):
         """Extract all elections."""
         # In most cases there isn't more than one 'Election' in a report, but the
         # standard allows more than one, so handle them.
@@ -282,7 +282,7 @@ def _extract_election_data(self, election_report: ElectionReport):
                 "start_date": election.start_date.strftime("%Y-%m-%d"),
                 "end_date": election.end_date.strftime("%Y-%m-%d"),
                 "ballot_styles": [
-                    _ for _ in self._extract_ballot_styles(election)
+                    _ for _ in self._election_ballot_styles(election)
                 ],
             }
             yield data
@@ -306,6 +306,6 @@ def extract(self, data: Dict, index: ElementIndex = None) -> ElectionData:
         self._index = index or ElementIndex(election_report, "ElectionResults")
         election_data = [
             ElectionData(**_)
-            for _ in self._extract_election_data(election_report)
+            for _ in self._elections(election_report)
         ]
         return election_data