itw-python-builder 0.2.14__tar.gz → 0.2.15__tar.gz

{itw_python_builder-0.2.14 → itw_python_builder-0.2.15}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: itw_python_builder
-Version: 0.2.14
+Version: 0.2.15
 Summary: Standardized Django deployment pipeline with Docker, testing, and SonarQube integration
 Author-email: IT-Works <contact@it-works.io>
 License: MIT
@@ -295,11 +295,11 @@ With `/repo`, leading and trailing slashes are stripped automatically, and full
 
 ### Milestones, Labels and Duplicates
 
-Milestones, labels and issue titles are matched with a shared fuzzy comparison at 95% similarity, so small differences in spacing or casing still count as a match. The threshold is a single constant (`SIMILARITY_THRESHOLD` in `task_utils.py`) if you need to tune it.
+Milestones, labels and issue titles are matched with a shared fuzzy comparison at 95% similarity, so small differences in spacing or casing still count as a match. Digit runs must be identical (`java 38` does not match `java 37`). The threshold is a single constant (`SIMILARITY_THRESHOLD` in `task_utils.py`) if you need to tune it.
 
-- **Milestone** — a similar existing milestone is reused; otherwise it is created (with the start/end dates when both are given).
+- **Milestone** — an existing project or parent-group milestone with the same title is reused; otherwise it is created (with the start/end dates when both are given).
 - **Label** — a similar existing label is linked and logged in white; a new one is created and logged in green. An empty `/label` is never sent to GitLab.
-- **Duplicate issues** — before creating, the target project is checked for a similar title *within the same milestone* (or among issues with no milestone). If one exists the issue is not created and a white log line reports it.
+- **Duplicate issues** — before creating, the target project is searched by title. A match in the same milestone, or an unassigned issue with the same title, is treated as a duplicate and skipped.
 
 ### Acceptance Criteria and Comments
 

{itw_python_builder-0.2.14 → itw_python_builder-0.2.15}/README.md

@@ -266,11 +266,11 @@ With `/repo`, leading and trailing slashes are stripped automatically, and full
 
 ### Milestones, Labels and Duplicates
 
-Milestones, labels and issue titles are matched with a shared fuzzy comparison at 95% similarity, so small differences in spacing or casing still count as a match. The threshold is a single constant (`SIMILARITY_THRESHOLD` in `task_utils.py`) if you need to tune it.
+Milestones, labels and issue titles are matched with a shared fuzzy comparison at 95% similarity, so small differences in spacing or casing still count as a match. Digit runs must be identical (`java 38` does not match `java 37`). The threshold is a single constant (`SIMILARITY_THRESHOLD` in `task_utils.py`) if you need to tune it.
 
-- **Milestone** — a similar existing milestone is reused; otherwise it is created (with the start/end dates when both are given).
+- **Milestone** — an existing project or parent-group milestone with the same title is reused; otherwise it is created (with the start/end dates when both are given).
 - **Label** — a similar existing label is linked and logged in white; a new one is created and logged in green. An empty `/label` is never sent to GitLab.
-- **Duplicate issues** — before creating, the target project is checked for a similar title *within the same milestone* (or among issues with no milestone). If one exists the issue is not created and a white log line reports it.
+- **Duplicate issues** — before creating, the target project is searched by title. A match in the same milestone, or an unassigned issue with the same title, is treated as a duplicate and skipped.
 
 ### Acceptance Criteria and Comments
 

{itw_python_builder-0.2.14 → itw_python_builder-0.2.15}/itw_python_builder/task_utils.py

@@ -1,9 +1,13 @@
 import re
+import unicodedata
 from difflib import SequenceMatcher
+from urllib.parse import quote
 
 from itw_python_builder.utils import _gitlab_api_call
 
 SIMILARITY_THRESHOLD = 0.95
+GITLAB_PAGE_SIZE = 100
+_RE_DIGITS = re.compile(r'\d+')
 
 DATE_PLACEHOLDER = 'YYYY-MM-DD'
 
@@ -48,7 +52,12 @@ def log_info(msg: str) -> None:
 
 
 def _normalize_title(value: str) -> str:
-    return ' '.join((value or '').lower().split())
+    collapsed = ' '.join((value or '').lower().split())
+    return unicodedata.normalize('NFC', collapsed)
+
+
+def _digit_tokens(value: str) -> list:
+    return _RE_DIGITS.findall(value or '')
 
 
 def is_similar(left: str, right: str, threshold: float = SIMILARITY_THRESHOLD) -> bool:
@@ -59,19 +68,64 @@ def is_similar(left: str, right: str, threshold: float = SIMILARITY_THRESHOLD) -
         return False
     if a == b:
         return True
+    if _digit_tokens(a) != _digit_tokens(b):
+        return False
     return SequenceMatcher(None, a, b).ratio() >= threshold
 
 
 def find_similar(value: str, items: list, key: str = 'title', threshold: float = SIMILARITY_THRESHOLD):
-    """Return the first item whose `key` is ~threshold similar to `value`, else None."""
+    """Return an exact normalized match if one exists, else the first ~threshold match."""
+    target = _normalize_title(value)
+    fuzzy = None
     for item in items:
-        if is_similar(value, item.get(key, ''), threshold):
+        other = item.get(key, '')
+        if target and target == _normalize_title(other):
             return item
-    return None
+        if fuzzy is None and is_similar(value, other, threshold):
+            fuzzy = item
+    return fuzzy
+
+
+def _gitlab_api_list(api_host: str, endpoint: str, token: str) -> list:
+    """GET a GitLab collection, following pages until a short page is returned."""
+    joiner = '&' if '?' in endpoint else '?'
+    page = 1
+    collected = []
+    while True:
+        chunk = _gitlab_api_call(
+            api_host,
+            f'{endpoint}{joiner}per_page={GITLAB_PAGE_SIZE}&page={page}',
+            token,
+        )
+        if not isinstance(chunk, list):
+            raise RuntimeError(
+                f'GitLab API GET {endpoint} expected a list, got {type(chunk).__name__}'
+            )
+        collected.extend(chunk)
+        if len(chunk) < GITLAB_PAGE_SIZE:
+            break
+        page += 1
+    return collected
+
+
+def _issue_milestone_title(issue: dict) -> str:
+    milestone = issue.get('milestone') or {}
+    if isinstance(milestone, dict):
+        return (milestone.get('title') or '').strip()
+    return ''
+
+
+def _is_duplicate_issue(issue: dict, title: str, milestone: str = None) -> bool:
+    """True when titles match and the issue is in the target milestone or has none."""
+    if not is_similar(title, issue.get('title', '')):
+        return False
+    existing = _issue_milestone_title(issue)
+    if milestone:
+        return (not existing) or is_similar(milestone, existing)
+    return not existing
 
 
 def _list_group_projects(api_host: str, group_path: str, token: str) -> list:
-    from urllib.parse import quote
     encoded = quote(group_path, safe='')
     return _gitlab_api_call(
         api_host,
@@ -113,27 +167,51 @@ def create_gitlab_issue(api_host: str, project_id: int, payload: dict, token: st
 
 
 def get_project_by_path(api_host: str, project_path: str, token: str) -> dict:
-    from urllib.parse import quote
     encoded = quote(project_path, safe='')
     return _gitlab_api_call(api_host, f'/projects/{encoded}', token)
 
 
 def find_project_milestone(api_host: str, project_id: int, title: str, token: str):
-    results = _gitlab_api_call(
+    """Find a project or ancestor-group milestone by title."""
+    encoded = quote(title, safe='')
+    base = f'/projects/{project_id}/milestones?include_ancestors=true'
+    exact = _gitlab_api_list(api_host, f'{base}&title={encoded}', token)
+    match = find_similar(title, exact)
+    if match:
+        return match
+    searched = _gitlab_api_list(api_host, f'{base}&search={encoded}', token)
+    return find_similar(title, searched)
+
+
+def _list_issues_for_milestone(api_host: str, project_id: int, milestone: str, token: str) -> list:
+    encoded = quote(milestone, safe='') if milestone else 'None'
+    return _gitlab_api_list(
         api_host,
-        f'/projects/{project_id}/milestones?per_page=100',
+        f'/projects/{project_id}/issues?milestone={encoded}',
         token,
     )
-    return find_similar(title, results)
 
 
 def find_project_issue(api_host: str, project_id: int, title: str, token: str, milestone: str = None):
-    """Look for a ~matching title within this milestone ('None' = the no-milestone bucket)."""
-    from urllib.parse import quote
-    endpoint = f'/projects/{project_id}/issues?per_page=100'
-    endpoint += f'&milestone={quote(milestone, safe="") if milestone else "None"}'
-    results = _gitlab_api_call(api_host, endpoint, token)
-    return find_similar(title, results)
+    """Look for a ~matching title in this milestone, or among issues with no milestone."""
+    encoded = quote(title, safe='')
+    searched = _gitlab_api_list(
+        api_host,
+        f'/projects/{project_id}/issues?search={encoded}&in=title',
+        token,
+    )
+    for item in searched:
+        if _is_duplicate_issue(item, title, milestone):
+            return item
+
+    for item in _list_issues_for_milestone(api_host, project_id, milestone, token):
+        if _is_duplicate_issue(item, title, milestone):
+            return item
+    if milestone:
+        for item in _list_issues_for_milestone(api_host, project_id, None, token):
+            if _is_duplicate_issue(item, title, milestone):
+                return item
+    return None
 
 
 def create_project_milestone(

{itw_python_builder-0.2.14 → itw_python_builder-0.2.15}/itw_python_builder.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: itw_python_builder
-Version: 0.2.14
+Version: 0.2.15
 Summary: Standardized Django deployment pipeline with Docker, testing, and SonarQube integration
 Author-email: IT-Works <contact@it-works.io>
 License: MIT
@@ -295,11 +295,11 @@ With `/repo`, leading and trailing slashes are stripped automatically, and full
 
 ### Milestones, Labels and Duplicates
 
-Milestones, labels and issue titles are matched with a shared fuzzy comparison at 95% similarity, so small differences in spacing or casing still count as a match. The threshold is a single constant (`SIMILARITY_THRESHOLD` in `task_utils.py`) if you need to tune it.
+Milestones, labels and issue titles are matched with a shared fuzzy comparison at 95% similarity, so small differences in spacing or casing still count as a match. Digit runs must be identical (`java 38` does not match `java 37`). The threshold is a single constant (`SIMILARITY_THRESHOLD` in `task_utils.py`) if you need to tune it.
 
-- **Milestone** — a similar existing milestone is reused; otherwise it is created (with the start/end dates when both are given).
+- **Milestone** — an existing project or parent-group milestone with the same title is reused; otherwise it is created (with the start/end dates when both are given).
 - **Label** — a similar existing label is linked and logged in white; a new one is created and logged in green. An empty `/label` is never sent to GitLab.
-- **Duplicate issues** — before creating, the target project is checked for a similar title *within the same milestone* (or among issues with no milestone). If one exists the issue is not created and a white log line reports it.
+- **Duplicate issues** — before creating, the target project is searched by title. A match in the same milestone, or an unassigned issue with the same title, is treated as a duplicate and skipped.
 
 ### Acceptance Criteria and Comments
 

{itw_python_builder-0.2.14 → itw_python_builder-0.2.15}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "itw_python_builder"
-version = "0.2.14"
+version = "0.2.15"
 description = "Standardized Django deployment pipeline with Docker, testing, and SonarQube integration"
 readme = "README.md"
 requires-python = ">=3.10"