fix setup.cfg

Tool now has round-trip accuracy with test file!

CHANGELOG.md

@@ -1,9 +1,24 @@
 # Changelog
 All notable changes to this project will be documented in this file
 
+## 0.2.0
+
+### Added
+
+- New yak log logo drawn by my sister
+
+### Changed
+
+- Updated package metadata
+- Rewrote parser to use a 2-step method that is more flexible.
+    - Parser can now handle code blocks.
+    - Parser can now handle setext-style headers and H2s not conforming to the
+      schema.
+
 ## 0.1.0 - 2021-04-16
 
 First release
 
 ### Added
+
 - `yaclog.read()` method to parse changelog files

README.md

@@ -1,2 +1,6 @@
-# yet-another-changelog
+# Yaclog
 Yet another changelog command line tool
+
+![a yak who is a log](https://github.com/drewcassidy/yaclog/raw/main/logo.png)
+
+*Logo by Erin Cassidy*

setup.cfg

@@ -7,7 +7,6 @@ license = AGPLv3
 license_file = LICENSE.md
 long_description = file: README.md
 long_description_content_type = text/markdown
-url = https://github.com/drewcassidy/yet-another-changelog
 
 keywords = changelog, commandline, markdown
 classifiers =
@@ -23,7 +22,11 @@ classifiers =
     Topic :: Software Development :: Version Control :: Git
     Topic :: Utilities
 
+project_urls =
+    Source Code = https://github.com/drewcassidy/yaclog
+    Changelog = https://github.com/drewcassidy/yaclog/blob/main/CHANGELOG.md
+
 [options]
 install_requires = Click; GitPython
-python_requires >= 3.8
+python_requires = >= 3.8
 packages = find:

yaclog/changelog.py

@@ -1,3 +1,19 @@
+#  yaclog: yet another changelog tool
+#  Copyright (c) 2021. Andrew Cassidy
+#
+#  This program is free software: you can redistribute it and/or modify
+#  it under the terms of the GNU Affero General Public License as
+#  published by the Free Software Foundation, either version 3 of the
+#  License, or (at your option) any later version.
+#
+#  This program is distributed in the hope that it will be useful,
+#  but WITHOUT ANY WARRANTY; without even the implied warranty of
+#  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+#  GNU Affero General Public License for more details.
+#
+#  You should have received a copy of the GNU Affero General Public License
+#  along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
 import datetime
 import os
 import re
@@ -7,6 +23,13 @@ from typing import List, Tuple, Optional
 bullets = '+-*'
 brackets = '[]'
 
+code_regex = re.compile(r'^```')
+header_regex = re.compile(r'^(?P<hashes>#+)\s+(?P<contents>[^#]+)(?:\s+#+)?$')
+under1_regex = re.compile(r'^=+\s*$')
+under2_regex = re.compile(r'^-+\s*$')
+bullet_regex = re.compile(r'^[-+*]')
+linkid_regex = re.compile(r'^\[(?P<link_id>\S*)]:\s*(?P<link>.*)')
+
 
 def _strip_link(token):
     if link_literal := re.fullmatch(r'\[(.*?)]\((.*?)\)', token):
@@ -20,6 +43,22 @@ def _strip_link(token):
     return token, None, None
 
 
+def _join_markdown(segments: List[str]) -> str:
+    text: List[str] = []
+    last_bullet = False
+    for segment in segments:
+        is_bullet = bullet_regex.match(segment) and '\n' not in segment
+
+        if not is_bullet or not last_bullet:
+            text.append('')
+
+        text.append(segment)
+
+        last_bullet = is_bullet
+
+    return '\n'.join(text).strip()
+
+
 class VersionEntry:
     def __init__(self):
         self.sections = {'': []}
@@ -27,17 +66,24 @@ class VersionEntry:
         self.date: Optional[datetime.date] = None
         self.tags: List[str] = []
         self.link: str = ''
+        self.link_id: str = None
+        self.line_no: int = -1
 
     def __str__(self) -> str:
-        if self.name.lower() == 'unreleased':
+        if self.link:
-            return f'## {self.name}'
+            segments = [f'[{self.name}]']
+        else:
+            segments = [self.name]
 
-        date_str = self.date.isoformat() if self.date else 'UNKNOWN'
+        if self.date or len(self.tags) > 0:
-        line = f'## {self.name} - {date_str}'
+            segments.append('-')
-        for tag in self.tags:
-            line += ' [' + tag.upper() + ']'
 
-        return line
+        if self.date:
+            segments.append(self.date.isoformat())
+
+        segments += [f'[{t.upper()}]' for t in self.tags]
+
+        return ' '.join(segments)
 
 
 class Changelog:
@@ -45,92 +91,168 @@ class Changelog:
         self.path = path
         self.header = ''
         self.versions = []
+
+        # Read file
+        with open(path, 'r') as fp:
+            self.lines = fp.readlines()
+
+        section = ''
+        in_block = False
+        in_code = False
+
         self.links = {}
 
-        with open(path, 'r') as fp:
+        links = {}
-            # Read file
+        segments: List[Tuple[int, List[str], str]] = []
-            line = fp.readline()
+        header_segments = []
-            while line and not line.startswith('##'):
-                self.header += line
-                line = fp.readline()
 
-            version = None
+        for line_no, line in enumerate(self.lines):
-            section = ''
+            if in_code:
-            last_line = ''
+                # this is the contents of a code block
+                segments[-1][1].append(line)
+                if code_regex.match(line):
+                    in_code = False
+                    in_block = False
 
-            while line:
+            elif code_regex.match(line):
-                if line.isspace():
+                # this is the start of a code block
-                    # skip empty lines
+                in_code = True
-                    pass
+                segments.append((line_no, [line], 'code'))
-                elif match := re.fullmatch(
-                        r'^##\s+(?P<name>\S*)(?:\s+-\s+(?P<date>\S+))?\s*?(?P<extra>.*?)\s*#*$', line):
-                    # this is a version header in the form '## Name (- date) (tags*) (#*)'
-                    version = VersionEntry()
-                    section = ''
 
-                    version.name, version.link, version.link_id = _strip_link(match['name'])
+            elif under1_regex.match(line) and in_block and len(segments[-1][1]) == 1 and segments[-1][2] == 'p':
+                # this is an underline for a setext-style H1
+                # ugly but it works
+                last = segments.pop()
+                segments.append((last[0], last[1] + [line], 'h1'))
 
-                    if match['date']:
+            elif under2_regex.match(line) and in_block and len(segments[-1][1]) == 1 and segments[-1][2] == 'p':
+                # this is an underline for a setext-style H2
+                # ugly but it works
+                last = segments.pop()
+                segments.append((last[0], last[1] + [line], 'h2'))
+
+            elif bullet_regex.match(line):
+                in_block = True
+                segments.append((line_no, [line], 'li'))
+
+            elif match := header_regex.match(line):
+                # this is a header
+                kind = f'h{len(match["hashes"])}'
+                segments.append((line_no, [line], kind))
+                in_block = False
+
+            elif match := linkid_regex.match(line):
+                # this is a link definition in the form '[id]: link', so add it to the link table
+                links[match['link_id'].lower()] = match['link']
+
+            elif line.isspace():
+                # skip empty lines
+                in_block = False
+
+            elif in_block:
+                # this is a line to be added to a paragraph
+                segments[-1][1].append(line)
+            else:
+                # this is a new paragraph
+                in_block = True
+                segments.append((line_no, [line], 'p'))
+
+        for segment in segments:
+            text = ''.join(segment[1]).strip()
+
+            if segment[2] == 'h2':
+                # start of a version
+
+                slug = text.rstrip('-').strip('#').strip()
+                split = slug.split()
+                if '-' in split:
+                    split.remove('-')
+
+                version = VersionEntry()
+                section = ''
+
+                version.name = slug
+                version.line_no = segment[0]
+                tags = []
+                date = []
+
+                for word in split[1:]:
+                    if match := re.match(r'\d{4}-\d{2}-\d{2}', word):
+                        # date
                         try:
-                            version.date = datetime.date.fromisoformat(match['date'].strip(string.punctuation))
+                            date = datetime.date.fromisoformat(match[0])
                         except ValueError:
-                            version.date = None
+                            break
-
+                    elif match := re.match(r'^\[(?P<tag>\S*)]', word):
-                    if match['extra']:
+                        tags.append(match['tag'])
-                        version.tags = [s.strip('[]') for s in re.findall(r'\[.*?]', match['extra'])]
+                    else:
-
+                        break
-                    self.versions.append(version)
-
-                elif match := re.fullmatch(r'###\s+(\S*?)(\s+#*)?', line):
-                    # this is a version section header in the form '### Name' or '### Name ###'
-                    section = match[1].title()
-                    if section not in version.sections.keys():
-                        version.sections[section] = []
-
-                elif match := re.fullmatch(r'\[(\S*)]:\s*(\S*)\n', line):
-                    # this is a link definition in the form '[id]: link', so add it to the link table
-                    self.links[match[1].lower()] = match[2]
-
-                elif line[0] in bullets or last_line.isspace():
-                    # bullet point or new paragraph
-                    # bullet points are preserved since some people like to use '+', '-' or '*' for different things
-                    version.sections[section].append(line.strip())
 
                 else:
-                    # not a bullet point, and no whitespace on last line, so append to the last entry
+                    # matches the schema
-                    version.sections[section][-1] += '\n' + line.strip()
+                    version.name, version.link, version.link_id = _strip_link(split[0])
+                    version.date = date
+                    version.tags = tags
 
-                last_line = line
+                self.versions.append(version)
-                line = fp.readline()
+
+            elif len(self.versions) == 0:
+                # we haven't encountered any version headers yet,
+                # so its best to just add this line to the header string
+                header_segments.append(text)
+
+            elif segment[2] == 'h3':
+                # start of a version section
+                section = text.strip('#').strip()
+                if section not in self.versions[-1].sections.keys():
+                    self.versions[-1].sections[section] = []
+
+            else:
+                # change log entry
+                self.versions[-1].sections[section].append(text)
+
+        # handle links
+        for version in self.versions:
+            if match := re.fullmatch(r'\[(.*)]', version.name):
+                # ref-matched link
+                link_id = match[1].lower()
+                if link_id in links:
+                    version.link = links.pop(link_id)
+                    version.link_id = None
+                    version.name = match[1]
+
+            elif version.link_id in links:
+                # id-matched link
+                version.link = links.pop(version.link_id)
+
+        # strip whitespace from header
+        self.header = _join_markdown(header_segments)
+        self.links = links
+
+    def write(self, path: os.PathLike = None):
+        if path is None:
+            path = self.path
+
+        v_links = {}
+        v_links.update(self.links)
+
+        with open(path, 'w') as fp:
+            fp.write(self.header)
+            fp.write('\n\n')
 
             for version in self.versions:
-                # handle links
+                fp.write(f'## {version}\n\n')
-                if match := re.fullmatch(r'\[(.*)]', version.name):
-                    # ref-matched link
-                    link_id = match[1].lower()
-                    if link_id in self.links:
-                        version.link = self.links.pop(link_id)
-                        version.link_id = None
-                        version.name = match[1]
 
-                elif version.link_id in self.links:
+                if version.link:
-                    # id-matched link
+                    v_links[version.name] = version.link
-                    version.link = self.links.pop(version.link_id)
 
+                for section in version.sections:
+                    if section:
+                        fp.write(f'### {section}\n\n')
 
-def read_version_header(line: str) -> Tuple[str, datetime.date, List[str]]:
+                    if len(version.sections[section]) > 0:
-    split = line.removeprefix('##').strip().split()
+                        fp.write(_join_markdown(version.sections[section]))
-    name = split[0]
+                        fp.write('\n\n')
-    date = datetime.date.fromisoformat(split[2]) if len(split) > 2 else None
-    tags = [s.removeprefix('[').removesuffix(']') for s in split[3:]]
 
-    return name, date, tags
+            for link_id, link in v_links.items():
-
+                fp.write(f'[{link_id.lower()}]: {link}\n')
-
-def write_version_header(name: str, date: datetime.date, tags=None) -> str:
-    line = f'## {name} - {date.isoformat()}'
-    if tags:
-        for tag in tags:
-            line += ' [' + tag.upper() + ']'
-
-    return line