fix(changelog): handle custom tag_format in changelog generation

When the tag_format does not follow the allowed schemas patterns then changlog generation fails.

Author: grahamhar

commitizen/changelog.py

@@ -93,16 +93,34 @@ def tag_included_in_changelog(
     return True
 
 
-def get_version_tags(scheme: type[BaseVersion], tags: list[GitTag]) -> list[GitTag]:
+def get_version_tags(
+    scheme: type[BaseVersion], tags: list[GitTag], tag_format: str
+) -> list[GitTag]:
     valid_tags: list[GitTag] = []
+    TAG_FORMAT_REGEXS = {
+        "$version": str(scheme.parser.pattern),
+        "$major": r"(?P<major>\d+)",
+        "$minor": r"(?P<minor>\d+)",
+        "$patch": r"(?P<patch>\d+)",
+        "$prerelease": r"(?P<prerelease>\w+\d+)?",
+        "$devrelease": r"(?P<devrelease>\.dev\d+)?",
+        "${version}": str(scheme.parser.pattern),
+        "${major}": r"(?P<major>\d+)",
+        "${minor}": r"(?P<minor>\d+)",
+        "${patch}": r"(?P<patch>\d+)",
+        "${prerelease}": r"(?P<prerelease>\w+\d+)?",
+        "${devrelease}": r"(?P<devrelease>\.dev\d+)?",
+    }
+    tag_format_regex = tag_format
+    for pattern, regex in TAG_FORMAT_REGEXS.items():
+        tag_format_regex = tag_format_regex.replace(pattern, regex)
     for tag in tags:
-        try:
-            scheme(tag.name)
-        except InvalidVersion:
-            out.warn(f"InvalidVersion {tag}")
-        else:
+        if re.match(tag_format_regex, tag.name):
             valid_tags.append(tag)
-
+        else:
+            out.warn(
+                f"InvalidVersion {tag.name} doesn't match configured tag format {tag_format}"
+            )
     return valid_tags
 
 
@@ -351,7 +369,6 @@ def get_oldest_and_newest_rev(
         oldest, newest = version.split("..")
     except ValueError:
         newest = version
-
     newest_tag = normalize_tag(newest, tag_format=tag_format, scheme=scheme)
 
     oldest_tag = None

commitizen/changelog_formats/base.py

@@ -25,10 +25,30 @@ def __init__(self, config: BaseConfig):
         # See: https://bugs.python.org/issue44807
         self.config = config
         self.encoding = self.config.settings["encoding"]
+        self.tag_format = self.config.settings.get("tag_format")
 
     @property
     def version_parser(self) -> Pattern:
-        return get_version_scheme(self.config).parser
+        version_regex = get_version_scheme(self.config).parser.pattern
+        if self.tag_format != "$version":
+            TAG_FORMAT_REGEXS = {
+                "$version": version_regex,
+                "$major": "(?P<major>\d+)",
+                "$minor": "(?P<minor>\d+)",
+                "$patch": "(?P<patch>\d+)",
+                "$prerelease": "(?P<prerelease>\w+\d+)?",
+                "$devrelease": "(?P<devrelease>\.dev\d+)?",
+                "${version}": version_regex,
+                "${major}": "(?P<major>\d+)",
+                "${minor}": "(?P<minor>\d+)",
+                "${patch}": "(?P<patch>\d+)",
+                "${prerelease}": "(?P<prerelease>\w+\d+)?",
+                "${devrelease}": "(?P<devrelease>\.dev\d+)?",
+            }
+            version_regex = self.tag_format
+            for pattern, regex in TAG_FORMAT_REGEXS.items():
+                version_regex = version_regex.replace(pattern, regex)
+        return rf"{version_regex}"
 
     def get_metadata(self, filepath: str) -> Metadata:
         if not os.path.isfile(filepath):

commitizen/commands/changelog.py

@@ -168,8 +168,10 @@ def __call__(self):
         # Don't continue if no `file_name` specified.
         assert self.file_name
 
-        tags = changelog.get_version_tags(self.scheme, git.get_tags()) or []
-
+        tags = (
+            changelog.get_version_tags(self.scheme, git.get_tags(), self.tag_format)
+            or []
+        )
         end_rev = ""
         if self.incremental:
             changelog_meta = self.changelog_format.get_metadata(self.file_name)
@@ -182,21 +184,18 @@ def __call__(self):
                 start_rev = self._find_incremental_rev(
                     strip_local_version(latest_tag_version), tags
                 )
-
         if self.rev_range:
             start_rev, end_rev = changelog.get_oldest_and_newest_rev(
                 tags,
                 version=self.rev_range,
                 tag_format=self.tag_format,
                 scheme=self.scheme,
             )
-
         commits = git.get_commits(start=start_rev, end=end_rev, args="--topo-order")
         if not commits and (
             self.current_version is None or not self.current_version.is_prerelease
         ):
             raise NoCommitsFoundError("No commits found")
-
         tree = changelog.generate_tree_from_commits(
             commits,
             tags,

commitizen/providers/scm_provider.py

@@ -29,6 +29,12 @@ class ScmProvider(VersionProvider):
         "$patch": r"(?P<patch>\d+)",
         "$prerelease": r"(?P<prerelease>\w+\d+)?",
         "$devrelease": r"(?P<devrelease>\.dev\d+)?",
+        "${version}": r"(?P<version>.+)",
+        "${major}": r"(?P<major>\d+)",
+        "${minor}": r"(?P<minor>\d+)",
+        "${patch}": r"(?P<patch>\d+)",
+        "${prerelease}": r"(?P<prerelease>\w+\d+)?",
+        "${devrelease}": r"(?P<devrelease>\.dev\d+)?",
     }
 
     def _tag_format_matcher(self) -> Callable[[str], VersionProtocol | None]:

docs/tutorials/monorepo_guidance.md

@@ -0,0 +1,40 @@
+# Configuring commitizen in a monorepo
+
+This tutorial assumes the monorepo layout is designed with multiple components that can be released independently of each other,
+some suggested layouts:
+
+```
+library-a
+  .cz.toml
+library-b
+  .cz.toml
+```
+
+```
+src
+    library-b
+      .cz.toml
+    library-z
+      .cz.toml
+```
+
+Each component will have its own changelog, commits will need to use scopes so only relevant commits are included in the
+appropriate change log for a given component. Example config and commit for `library-b`
+
+```toml
+[tool.commitizen]
+name = "cz_customize"
+version = "0.0.0"
+tag_format = "${version}-library-b" # the component name can be a prefix or suffix with or without a separator
+update_changelog_on_bump = true
+
+[tool.commitizen.customize]
+changelog_pattern = "^(feat|fix)\\(library-b\\)(!)?:" #the pattern on types can be a wild card or any types you wish to include
+```
+
+example commit message for the above
+
+`fix:(library-b) Some awesome message`
+
+If the above is followed and the `cz bump --changelog` is run in the directory containing the component the changelog
+should be generated in the same directory with only commits scoped to the component.