Fix; detect wrong title level for changes & handle duplicate sections (#5)

Author: mdwint

CHANGELOG.md

@@ -6,6 +6,15 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
 
+## 0.2.2 - 2022-08-01
+
+### Fixed
+
+- Raise error on wrong title level for changes.
+- Raise error on duplicate version sections.
+- Merge duplicate changes sections.
+
+
 ## 0.2.1 - 2022-08-01
 
 ### Fixed

ocdc/__init__.py

@@ -1 +1 @@
-__version__ = "0.2.1"
+__version__ = "0.2.2"

ocdc/ast.py

@@ -22,6 +22,10 @@ class Changes(BaseModel):
     items: List[ListItem]
     footer: str = ""
 
+    def merge(self, other: "Changes") -> None:
+        self.items.extend(other.items)
+        self.footer += "\n" + other.footer
+
 
 class Version(BaseModel):
     number: str

ocdc/parser.py

@@ -1,6 +1,7 @@
+from contextlib import contextmanager
 from dataclasses import dataclass
 from enum import Enum, auto
-from typing import Any, List, Optional, Set, Tuple
+from typing import Any, Dict, List, Optional, Set, Tuple
 
 from . import ast
 from .renderer import INDENT
@@ -12,15 +13,35 @@ def parse(text: str) -> ast.Changelog:
 
 
 class ParseError(Exception):
-    def __init__(self, msg: str, source: str, row: int, col_start: int, col_end: int):
-        lines = source.splitlines()
-        prev, line = lines[row - 1 : row + 1] if row > 0 else ("", lines[row])
+    def __init__(
+        self,
+        msg: str,
+        source: str,
+        token: "Token",
+        back: int = 0,
+        forward: int = 0,
+        hint: str = "",
+    ):
+        msg = f"{msg} {annotate(source, token, back, forward)}"
+        if hint:
+            msg += f"\n\nHint: {hint}"
+        super().__init__(msg)
 
-        arrows = " " * col_start + "^" * (col_end - col_start)
-        details = (f"  {prev}\n" if prev else "") + f"  {line}\n  {arrows}"
 
-        msg = f"{msg} (line {row + 1}, column {col_start + 1}):\n\n{details}"
-        super().__init__(msg)
+def annotate(source: str, token: "Token", back: int = 0, forward: int = 0) -> str:
+    lines = source.splitlines()
+    row = min(token.row, len(lines) - 1)
+    prev, line = lines[row - 1 : row + 1] if row > 0 else ("", lines[row])
+
+    col_start = col_end = token.col
+    if back:
+        col_start -= back
+    if forward:
+        col_end += forward
+
+    arrows = " " * col_start + "^" * (col_end - col_start)
+    details = (f"  {prev}\n" if prev else "") + f"  {line}\n  {arrows}"
+    return f"at line {row + 1}, column {col_start + 1}:\n\n{details}"
 
 
 class TokenType(Enum):
@@ -35,7 +56,6 @@ class TokenType(Enum):
 @dataclass
 class Token:
     typ: TokenType
-    pos: int
     row: int
     col: int
     text: str
@@ -70,7 +90,7 @@ def peek(self, offset: int = 0) -> str:
 
     def add_token(self, typ: TokenType, text: str = "", parsed: Any = None):
         col = self.col - len(self.matched)
-        token = Token(typ, self.start, self.row, col, text or self.matched, parsed)
+        token = Token(typ, self.row, col, text or self.matched, parsed)
         self.tokens.append(token)
 
     def __call__(self) -> List[Token]:
@@ -114,6 +134,17 @@ def __init__(self, source: str, tokens: List[Token]):
     def has_more(self) -> bool:
         return self.peek().typ != TokenType.EOF
 
+    class Rollback(Exception):
+        pass
+
+    @contextmanager
+    def checkpoint(self):
+        saved = self.current
+        try:
+            yield
+        except self.Rollback:
+            self.current = saved
+
     def advance(self, n: int = 1):
         if self.has_more:
             self.current += n
@@ -132,6 +163,12 @@ def match(self, types: Set[TokenType]) -> bool:
                 return True
         return False
 
+    def match_many(self, typ: TokenType) -> int:
+        n = 0
+        while self.match({typ}):
+            n += 1
+        return n
+
     def expect(self, *types: TokenType) -> bool:
         n = len(types)
         if self.current + n < len(self.tokens):
@@ -141,15 +178,15 @@ def expect(self, *types: TokenType) -> bool:
         return False
 
     def error(
-        self, msg: str, token: Optional[Token] = None, back: int = 0, forward: int = 0
+        self,
+        msg: str,
+        token: Optional[Token] = None,
+        back: int = 0,
+        forward: int = 0,
+        hint: str = "",
     ) -> ParseError:
         token = token or self.peek()
-        col_start = col_end = token.col
-        if back:
-            col_start -= back
-        if forward:
-            col_end += forward
-        return ParseError(msg, self.source, token.row, col_start, col_end)
+        return ParseError(msg, self.source, token, back, forward, hint)
 
     def __call__(self) -> ast.Changelog:
         return changelog(self)
@@ -167,8 +204,19 @@ def changelog(p: Parser) -> ast.Changelog:
         while p.match({TokenType.FOOTER_BEGIN}):
             c.intro += "\n" + text(p)
 
+    seen_version_titles: Dict[str, Token] = {}
     while p.expect(TokenType.HASH, TokenType.HASH, TokenType.TEXT):
-        c.versions.append(version(p))
+        v, title_token = version(p)
+
+        duplicate_title_token = seen_version_titles.get(v.number)
+        if duplicate_title_token:
+            forward = len(v.number)
+            at = annotate(p.source, title_token, forward=forward)
+            msg, hint = "Duplicate version", f"Version was first used {at}"
+            raise p.error(msg, duplicate_title_token, forward=forward, hint=hint)
+
+        seen_version_titles[v.number] = title_token
+        c.versions.append(v)
 
     if p.has_more:
         raise p.error("Unprocessable text", forward=len(p.peek().text))
@@ -183,8 +231,9 @@ def text(p: Parser) -> str:
     return text.strip()
 
 
-def version(p: Parser) -> ast.Version:
-    title = p.peek(1).text
+def version(p: Parser) -> Tuple[ast.Version, Token]:
+    title_token = p.peek(1)
+    title = title_token.text
     try:
         number, date = title.split(" - ", 1)
     except ValueError:
@@ -196,9 +245,28 @@ def version(p: Parser) -> ast.Version:
 
     while p.expect(TokenType.HASH, TokenType.HASH, TokenType.HASH, TokenType.TEXT):
         type_, changes_ = changes(p)
-        v.changes[type_.value] = changes_
+        key = type_.value
+        if key in v.changes:
+            v.changes[key].merge(changes_)
+        else:
+            v.changes[key] = changes_
+
+    detect_wrong_title_level(p, 3)
+
+    return v, title_token
+
 
-    return v
+def detect_wrong_title_level(p: Parser, expected: int) -> None:
+    with p.checkpoint():
+        level = p.match_many(TokenType.HASH)
+        if level and level != expected:
+            token = p.peek(1)
+            p.match({TokenType.TEXT})
+            skip_newlines(p)
+            if p.match({TokenType.DASH}):
+                msg = f"Expected a title of H{expected}, but found H{level}"
+                raise p.error(msg, token, back=level)
+        raise p.Rollback
 
 
 def changes(p: Parser) -> Tuple[ast.ChangeType, ast.Changes]:
@@ -207,8 +275,9 @@ def changes(p: Parser) -> Tuple[ast.ChangeType, ast.Changes]:
     try:
         type_ = getattr(ast.ChangeType, title)
     except AttributeError:
-        err = p.error("Unexpected title for changes", token, forward=len(title))
-        raise err from None
+        msg = "Unexpected title for changes"
+        hint = "Choose from " + ", ".join(f'"{t.value}"' for t in ast.ChangeType) + "."
+        raise p.error(msg, token, forward=len(title), hint=hint) from None
 
     skip_newlines(p)
 

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "ocdc"
-version = "0.2.1"
+version = "0.2.2"
 description = 'A changelog formatter for "people", neat freaks, and sloppy typists.'
 authors = ["Matteo De Wint <[email protected]>"]
 packages = [{ include = "ocdc" }]

setup.cfg

@@ -1,5 +1,5 @@
 [bumpversion]
-current_version = 0.2.1
+current_version = 0.2.2
 parse = (?P<major>\d+)\.(?P<minor>\d+)\.(?P<patch>\d+)
 serialize = 
 	{major}.{minor}.{patch}

tests/data/duplicate_changes_sections.ast.json

@@ -0,0 +1,30 @@
+{
+  "title": "Changelog",
+  "intro": "All notable changes to this project will be documented in this file.\n\nThe format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),\nand this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).",
+  "versions": [
+    {
+      "number": "0.1.0",
+      "date": "2022-07-31",
+      "changes": {
+        "Added": {
+          "items": [
+            {
+              "text": "Something in [one]."
+            },
+            {
+              "text": "Something in [two]."
+            }
+          ],
+          "footer": "[one]: http://example.com/one\n[two]: http://example.com/two"
+        },
+        "Fixed": {
+          "items": [
+            {
+              "text": "A bug."
+            }
+          ]
+        }
+      }
+    }
+  ]
+}

tests/data/duplicate_changes_sections.in.md

@@ -0,0 +1,24 @@
+# Changelog
+
+All notable changes to this project will be documented in this file.
+
+The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
+and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
+
+## 0.1.0 - 2022-07-31
+
+### Added
+
+- Something in [one].
+
+[one]: http://example.com/one
+
+### Fixed
+
+- A bug.
+
+### Added
+
+- Something in [two].
+
+[two]: http://example.com/two

tests/data/duplicate_changes_sections.out.md

@@ -0,0 +1,21 @@
+# Changelog
+
+All notable changes to this project will be documented in this file.
+
+The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
+and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
+
+
+## 0.1.0 - 2022-07-31
+
+### Added
+
+- Something in [one].
+- Something in [two].
+
+[one]: http://example.com/one
+[two]: http://example.com/two
+
+### Fixed
+
+- A bug.

tests/data/duplicate_version_sections.err.md

@@ -0,0 +1,24 @@
+# Changelog
+
+All notable changes to this project will be documented in this file.
+
+The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
+and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
+
+## 0.1.0 - 2022-08-02
+
+### Added
+
+- Something else.
+
+## 0.2.0 - 2022-08-01
+
+### Added
+
+- Something.
+
+## 0.1.0 - 2022-07-31
+
+### Added
+
+- Initial version.