Jira issue provider support (minimal)

docs/docs/core-abilities/fetching_ticket_context.md

@@ -164,6 +164,16 @@ jira_api_token = "YOUR_API_TOKEN"
 jira_api_email = "YOUR_EMAIL"
 ```
 
+To use Jira as the issue provider for ticket compliance (and `/similar_issue`), enable it explicitly:
+
+```toml
+[config]
+issue_provider = "jira"
+
+[jira]
+issue_projects = ["ABC"] # or issue_jql = "project = ABC order by created DESC"
+```
+
 ### Jira Data Center/Server
 
 [//]: # ()

docs/docs/tools/similar_issues.md

@@ -1,7 +1,7 @@
 ## Overview
 
-The similar issue tool retrieves the most similar issues to the current issue.
-It can be invoked manually by commenting on any PR:
+The similar issue tool retrieves the most similar issues to the current issue or MR context.
+It can be invoked manually by commenting on any PR/MR:
 
 ```
 /similar_issue
@@ -15,8 +15,69 @@ It can be invoked manually by commenting on any PR:
 
 ![similar_issue](https://codium.ai/images/pr_agent/similar_issue.png){width=768}
 
+### GitLab example (MR comment)
+
+Comment on an MR:
+
+```
+/similar_issue
+```
+
+Example output posted to the MR:
+
+```
+### Similar Issues
+___
+
+1. **[Add retry logic for HTTP client](https://gitlab.example.com/org/repo/-/issues/1)** (score=0.91)
+2. **[Cache embeddings for faster review](https://gitlab.example.com/org/repo/-/issues/3)** (score=0.89)
+```
+
 Note that to perform retrieval, the `similar_issue` tool indexes all the repo previous issues (once).
 
+## Indexing lifecycle and scope
+
+### What is indexed
+- Issues and (optionally) issue comments only. MRs are not indexed.
+- Each vector includes `repo`, `username`, `created_at`, and `level` (issue or comment).
+
+### When indexing happens
+- On demand, the first time `/similar_issue` is called for a repo.
+- A per-repo marker record is stored to avoid re-indexing the same repo.
+- On later runs, only new issues are appended (based on issue IDs).
+
+### Query scope
+- One shared collection is used, but queries always filter to the current repo.
+- GitLab: the query text comes from MR title + description. If the MR text includes `#<issue>`, that GitLab issue is used as the query source, but the output still posts on the MR.
+
+```mermaid
+flowchart TD
+    A[Comment /similar_issue on MR] --> B{Repo indexed?}
+    B -- No --> C[Fetch repo issues + comments]
+    C --> D[Embed + upsert vectors to vector DB]
+    B -- Yes --> E[Check for new issues]
+    E --> F{New issues?}
+    F -- Yes --> D
+    F -- No --> G[Build query]
+    D --> G[Build query]
+    G --> H[Query vector DB (filter by repo)]
+    H --> I[Post Similar Issues on MR]
+```
+
+## Embedding configuration
+
+The tool uses an OpenAI-compatible embeddings endpoint. Configure it in `configuration.toml` (or via env vars):
+
+```
+[pr_similar_issue]
+embedding_base_url = "https://your-embeddings-host/v1/embeddings"
+embedding_model = "intfloat/multilingual-e5-large"
+embedding_dim = 1024
+embedding_max_tokens = 10000
+```
+
+If the embedding endpoint requires auth, set `PR_SIMILAR_ISSUE__EMBEDDING_API_KEY` as an environment variable.
+
 ### Selecting a Vector Database
 
 Configure your preferred database by changing the `pr_similar_issue` parameter in `configuration.toml` file.
@@ -59,13 +120,18 @@ vectordb = "qdrant"
 ```
 
 You can get a free managed Qdrant instance from [Qdrant Cloud](https://cloud.qdrant.io/).
+Ensure the Qdrant collection dimension matches `embedding_dim`. If you change models, set
+`pr_similar_issue.force_update_dataset=true` to rebuild the collection.
 
 ## How to use
 
 - To invoke the 'similar issue' tool from **CLI**, run:
 `python3 cli.py --issue_url=... similar_issue`
 
-- To invoke the 'similar' issue tool via online usage, [comment](https://github.com/Codium-ai/pr-agent/issues/178#issuecomment-1716934893) on a PR:
+- To invoke the 'similar issue' tool via online usage, [comment](https://github.com/Codium-ai/pr-agent/issues/178#issuecomment-1716934893) on a PR/MR:
 `/similar_issue`
 
+- GitLab: if run from an MR comment, the query uses the MR title + description. If the MR text includes an issue reference (e.g., `#123`), that issue is used as the query source, but the output is still posted on the MR. If run from CLI with `--issue_url`, the query uses that issue.
+- Jira: set `issue_provider="jira"` and configure `[jira]` with either `issue_projects` (or `issue_project_map`) or `issue_jql`. When enabled, `/similar_issue` indexes Jira issues instead of GitLab/GitHub issues. If the MR text includes Jira keys (e.g., `ABC-123`), those tickets are used as the query source; otherwise it uses the MR title + description.
+
 - You can also enable the 'similar issue' tool to run automatically when a new issue is opened, by adding it to the [pr_commands list in the github_app section](https://github.com/Codium-ai/pr-agent/blob/main/pr_agent/settings/configuration.toml#L66)

pr_agent/algo/ticket_utils.py

@@ -0,0 +1,16 @@
+import re
+from typing import List
+
+JIRA_KEY_PATTERN = re.compile(r"(?:https?://[^\s/]+/browse/)?([A-Z][A-Z0-9]+-\d{1,7})", re.IGNORECASE)
+
+
+def find_jira_keys(text: str) -> List[str]:
+    if not text:
+        return []
+    matches = JIRA_KEY_PATTERN.findall(text)
+    keys = []
+    for match in matches:
+        key = match.upper()
+        if key not in keys:
+            keys.append(key)
+    return keys

pr_agent/git_providers/gitlab_provider.py

@@ -71,7 +71,8 @@ def __init__(self, merge_request_url: Optional[str] = None, incremental: Optiona
         self.temp_comments = []
         self._submodule_cache: dict[tuple[str, str, str], list[dict]] = {}
         self.pr_url = merge_request_url
-        self._set_merge_request(merge_request_url)
+        if merge_request_url and self._is_merge_request_url(merge_request_url):
+            self._set_merge_request(merge_request_url)
         self.RE_HUNK_HEADER = re.compile(
             r"^@@ -(\d+)(?:,(\d+))? \+(\d+)(?:,(\d+))? @@[ ]?(.*)")
         self.incremental = incremental
@@ -785,9 +786,6 @@ def get_pr_owner_id(self) -> str | None:
     def get_pr_description_full(self):
         return self.mr.description
 
-    def get_issue_comments(self):
-        return self.mr.notes.list(get_all=True)[::-1]
-
     def get_repo_settings(self):
         try:
             main_branch = self.gl.projects.get(self.id_project).default_branch
@@ -847,6 +845,13 @@ def remove_reaction(self, issue_comment_id: int, reaction_id: str) -> bool:
             get_logger().warning(f"Failed to remove reaction, error: {e}")
             return False
 
+    def _is_merge_request_url(self, url: str) -> bool:
+        try:
+            path_parts = urlparse(url).path.strip('/').split('/')
+        except Exception:
+            return False
+        return "merge_requests" in path_parts
+
     def _parse_merge_request_url(self, merge_request_url: str) -> Tuple[str, int]:
         parsed_url = urlparse(merge_request_url)
 
@@ -872,10 +877,64 @@ def _parse_merge_request_url(self, merge_request_url: str) -> Tuple[str, int]:
         # Return the path before 'merge_requests' and the ID
         return project_path, mr_id
 
+    def _parse_issue_url(self, issue_url: str) -> Tuple[str, int]:
+        parsed_url = urlparse(issue_url)
+
+        path_parts = parsed_url.path.strip('/').split('/')
+        if 'issues' not in path_parts:
+            raise ValueError("The provided URL does not appear to be a GitLab issue URL")
+
+        issues_index = path_parts.index('issues')
+        if len(path_parts) <= issues_index + 1:
+            raise ValueError("The provided URL does not contain an issue IID")
+
+        try:
+            issue_iid = int(path_parts[issues_index + 1])
+        except ValueError as e:
+            raise ValueError("Unable to convert issue IID to integer") from e
+
+        project_parts = path_parts[:issues_index]
+        if project_parts and project_parts[-1] == '-':
+            project_parts = project_parts[:-1]
+        project_path = "/".join(project_parts)
+        if project_path.endswith('/-'):
+            project_path = project_path[:-2]
+        return project_path, issue_iid
+
     def _get_merge_request(self):
         mr = self.gl.projects.get(self.id_project).mergerequests.get(self.id_mr)
         return mr
 
+    def _get_project(self, project_path: str):
+        try:
+            encoded = urllib.parse.quote_plus(project_path)
+            return self.gl.projects.get(encoded)
+        except Exception:
+            return self._project_by_path(project_path)
+
+    def get_issue(self, issue_iid: int, project_path: Optional[str] = None):
+        project = self._get_project(project_path or self.id_project)
+        if project is None:
+            raise GitlabGetError("Project not found")
+        return project.issues.get(issue_iid)
+
+    def list_issues(self, project_path: Optional[str] = None, state: str = "all"):
+        project = self._get_project(project_path or self.id_project)
+        if project is None:
+            raise GitlabGetError("Project not found")
+        return project.issues.list(state=state, iterator=True)
+
+    def get_issue_comments(self, issue=None):
+        if issue is None:
+            try:
+                return self.mr.notes.list(get_all=True)[::-1]
+            except Exception:
+                return []
+        return list(issue.notes.list(iterator=True))
+
+    def create_issue_comment(self, issue, body: str):
+        return issue.notes.create({"body": body})
+
     def get_user_id(self):
         return None
 
@@ -954,22 +1013,36 @@ def generate_link_to_relevant_line_number(self, suggestion) -> str:
         return ""
     #Clone related
     def _prepare_clone_url_with_token(self, repo_url_to_clone: str) -> str | None:
-        if "gitlab." not in repo_url_to_clone:
-            get_logger().error(f"Repo URL: {repo_url_to_clone} is not a valid gitlab URL.")
-            return None
-        (scheme, base_url) = repo_url_to_clone.split("gitlab.")
         access_token = getattr(self.gl, 'oauth_token', None) or getattr(self.gl, 'private_token', None)
-        if not all([scheme, access_token, base_url]):
-            get_logger().error(f"Either no access token found, or repo URL: {repo_url_to_clone} "
-                               f"is missing prefix: {scheme} and/or base URL: {base_url}.")
+        if not access_token:
+            get_logger().error("No access token found for GitLab clone.")
             return None
 
-        #Note that the ""official"" method found here:
-        # https://docs.gitlab.com/user/profile/personal_access_tokens/#clone-repository-using-personal-access-token
-        # requires a username, which may not be applicable.
-        # The following solution is taken from: https://stackoverflow.com/questions/25409700/using-gitlab-token-to-clone-without-authentication/35003812#35003812
-        # For example: For repo url: https://gitlab.codium-inc.com/qodo/autoscraper.git
-        # Then to clone one will issue: 'git clone https://oauth2:<access token>@gitlab.codium-inc.com/qodo/autoscraper.git'
+        # Note: GitLab instances are not always hosted under a gitlab.* domain.
+        # Build a clone URL that works with any host (e.g., git.labs.hosting.cerence.net).
+        if repo_url_to_clone.startswith(("http://", "https://")):
+            try:
+                from urllib.parse import urlparse
+                parsed = urlparse(repo_url_to_clone)
+                if not parsed.scheme or not parsed.netloc:
+                    raise ValueError("missing scheme or host")
+                netloc = parsed.netloc.split("@")[-1]
+                return f"{parsed.scheme}://oauth2:{access_token}@{netloc}{parsed.path}"
+            except Exception as exc:
+                get_logger().error(
+                    f"Repo URL: {repo_url_to_clone} could not be parsed for clone.",
+                    artifact={"error": str(exc)},
+                )
+                return None
 
-        clone_url = f"{scheme}oauth2:{access_token}@gitlab.{base_url}"
-        return clone_url
+        # Fallback to legacy gitlab.* parsing when a raw URL is provided.
+        if "gitlab." not in repo_url_to_clone:
+            get_logger().error(f"Repo URL: {repo_url_to_clone} is not a valid gitlab URL.")
+            return None
+        scheme, base_url = repo_url_to_clone.split("gitlab.")
+        if not all([scheme, base_url]):
+            get_logger().error(
+                f"Repo URL: {repo_url_to_clone} is missing prefix: {scheme} and/or base URL: {base_url}."
+            )
+            return None
+        return f"{scheme}oauth2:{access_token}@gitlab.{base_url}"

pr_agent/issue_providers/__init__.py

@@ -0,0 +1,16 @@
+from pr_agent.issue_providers.base import Issue, IssueComment, IssueProvider
+from pr_agent.issue_providers.github_issue_provider import GithubIssueProvider
+from pr_agent.issue_providers.gitlab_issue_provider import GitlabIssueProvider
+from pr_agent.issue_providers.jira_issue_provider import JiraIssueProvider
+from pr_agent.issue_providers.resolver import get_issue_provider, resolve_issue_provider_name
+
+__all__ = [
+    "Issue",
+    "IssueComment",
+    "IssueProvider",
+    "GithubIssueProvider",
+    "GitlabIssueProvider",
+    "JiraIssueProvider",
+    "get_issue_provider",
+    "resolve_issue_provider_name",
+]

pr_agent/issue_providers/base.py

@@ -0,0 +1,46 @@
+from __future__ import annotations
+
+from abc import ABC, abstractmethod
+from dataclasses import dataclass, field
+from typing import Iterable, List, Optional
+
+
+@dataclass
+class IssueComment:
+    body: str
+    url: str = ""
+    id: Optional[str] = None
+    author: Optional[str] = None
+
+
+@dataclass
+class Issue:
+    key: str
+    title: str
+    description: str = ""
+    url: str = ""
+    created_at: Optional[str] = None
+    author: Optional[dict] = None
+    comments: List[IssueComment] = field(default_factory=list)
+    labels: List[str] = field(default_factory=list)
+
+    @property
+    def body(self) -> str:
+        return self.description
+
+    @property
+    def web_url(self) -> str:
+        return self.url
+
+
+class IssueProvider(ABC):
+    @abstractmethod
+    def list_issues(self, project_path: Optional[str] = None, state: str = "all") -> Iterable:
+        raise NotImplementedError
+
+    @abstractmethod
+    def get_issue(self, issue_id: str, project_path: Optional[str] = None):
+        raise NotImplementedError
+
+    def get_issue_comments(self, issue) -> List[IssueComment]:
+        return []

pr_agent/issue_providers/github_issue_provider.py

@@ -0,0 +1,21 @@
+from __future__ import annotations
+
+from typing import Optional
+
+from pr_agent.issue_providers.base import IssueProvider
+
+
+class GithubIssueProvider(IssueProvider):
+    def __init__(self, git_provider, repo_obj):
+        self.git_provider = git_provider
+        self.repo_obj = repo_obj
+
+    def list_issues(self, project_path: Optional[str] = None, state: str = "all"):
+        return self.repo_obj.get_issues(state=state)
+
+    def get_issue(self, issue_id, project_path: Optional[str] = None):
+        issue_number = int(issue_id)
+        return self.repo_obj.get_issue(issue_number)
+
+    def get_issue_comments(self, issue):
+        return list(issue.get_comments())

pr_agent/issue_providers/gitlab_issue_provider.py

@@ -0,0 +1,20 @@
+from __future__ import annotations
+
+from typing import Optional
+
+from pr_agent.issue_providers.base import IssueProvider
+
+
+class GitlabIssueProvider(IssueProvider):
+    def __init__(self, git_provider):
+        self.git_provider = git_provider
+
+    def list_issues(self, project_path: Optional[str] = None, state: str = "all"):
+        return self.git_provider.list_issues(project_path, state=state)
+
+    def get_issue(self, issue_id, project_path: Optional[str] = None):
+        issue_iid = int(issue_id)
+        return self.git_provider.get_issue(issue_iid, project_path)
+
+    def get_issue_comments(self, issue):
+        return self.git_provider.get_issue_comments(issue)