Ran pre-commit on cevatkerim's branch and fixed type hints to ensure CI passes

README.md

@@ -19,6 +19,9 @@ You can also replace `hub` with `ingest` in any github url to access the corespo
 
 - **Easy code context**: Get a text digest from a git repository URL or a directory
 - **Smart Formatting**: Optimized output format for LLM prompts
+- **Flexible Filtering**:
+  - Include/exclude files by pattern
+  - Control maximum file size
 - **Statistics about**:
   - File and directory structure
   - Size of the extract
@@ -62,6 +65,31 @@ summary, tree, content = ingest("https://github.com/cyclotruc/gitingest")
 
 By default, this won't write a file but can be enabled with the `output` argument
 
+## 🌐 Web API Usage
+
+You can access repositories directly via URL:
+
+```plaintext
+# Basic repository access
+https://gitingest.com/user/repo
+
+# With query parameters
+https://gitingest.com/extract/user/repo?max_file_size=243&pattern_type=include&pattern=*.py
+
+# Parameters:
+- max_file_size: Controls maximum file size (default: 243)
+- pattern_type: 'include' or 'exclude'
+- pattern: File pattern (e.g. "*.py" for Python files)
+- summary: Set to true to include summary and tree structure
+```
+
+## 🛠️ Using
+
+- Tailwind CSS - Frontend
+- [FastAPI](https://github.com/fastapi/fastapi) - Backend framework
+- [tiktoken](https://github.com/openai/tiktoken) - Token estimation
+- [apianalytics.dev](https://www.apianalytics.dev/) - Simple Analytics
+
 ## 🌐 Self-host
 
 1. Build the image:

src/process_query.py

@@ -18,40 +18,43 @@ async def process_query(
     pattern_type: str = "exclude",
     pattern: str = "",
     is_index: bool = False,
-) -> _TemplateResponse:
+    raw_response: bool = False,
+) -> _TemplateResponse | tuple[str, str, str]:
     """
-    Process a query by parsing input, cloning a repository, and generating a summary.
-
-    Handle user input, process GitHub repository data, and prepare
-    a response for rendering a template with the processed results or an error message.
+    Process query and return template response or raw data tuple.
 
     Parameters
     ----------
     request : Request
-        The HTTP request object.
+    HTTP request object
     input_text : str
-        Input text provided by the user, typically a GitHub repository URL or slug.
+    GitHub repository URL or slug
     slider_position : int
-        Position of the slider, representing the maximum file size in the query.
+    Maximum file size position (0-500)
     pattern_type : str, optional
-        Type of pattern to use, either "include" or "exclude" (default is "exclude").
+    "include" or "exclude" pattern type (default: "exclude")
     pattern : str, optional
-        Pattern to include or exclude in the query, depending on the pattern type.
+    Pattern for including/excluding files
     is_index : bool, optional
-        Flag indicating whether the request is for the index page (default is False).
+    Whether request is for index page (default: False)
+    raw_response : bool, optional
+    Return raw data tuple instead of template (default: False)
 
     Returns
     -------
-    _TemplateResponse
-        Rendered template response containing the processed results or an error message.
+    Union[_TemplateResponse, tuple[str, str, str]]
+    TemplateResponse:
+        Rendered HTML template with processed results, summary, and error messages
+    tuple[str, str, str]:
+        Raw data as (summary, directory_tree, file_contents) when raw_response=True
     """
     template = "index.jinja" if is_index else "github.jinja"
     max_file_size = logSliderToSize(slider_position)
 
     if pattern_type == "include":
         include_patterns = pattern
         exclude_patterns = None
-    elif pattern_type == "exclude":
+    else:
         exclude_patterns = pattern
         include_patterns = None
 
@@ -63,17 +66,53 @@ async def process_query(
             include_patterns=include_patterns,
             ignore_patterns=exclude_patterns,
         )
+
         clone_config = CloneConfig(
             url=query["url"],
             local_path=query["local_path"],
             commit=query.get("commit"),
             branch=query.get("branch"),
         )
+
         await clone_repo(clone_config)
         summary, tree, content = ingest_from_query(query)
+
+        if raw_response:
+            return summary, tree, content
+
         with open(f"{clone_config.local_path}.txt", "w") as f:
             f.write(tree + "\n" + content)
 
+        if not raw_response and len(content) > MAX_DISPLAY_SIZE:
+            content = (
+                f"(Files content cropped to {int(MAX_DISPLAY_SIZE / 1_000)}k characters, "
+                "download full ingest to see more)\n" + content[:MAX_DISPLAY_SIZE]
+            )
+
+        _print_success(
+            url=query["url"],
+            max_file_size=max_file_size,
+            pattern_type=pattern_type,
+            pattern=pattern,
+            summary=summary,
+        )
+        return templates.TemplateResponse(
+            template,
+            {
+                "request": request,
+                "github_url": input_text,
+                "result": True,
+                "summary": summary,
+                "tree": tree,
+                "content": content,
+                "examples": EXAMPLE_REPOS if is_index else [],
+                "ingest_id": query["id"],
+                "default_file_size": slider_position,
+                "pattern_type": pattern_type,
+                "pattern": pattern,
+            },
+        )
+
     except Exception as e:
         # hack to print error message when query is not defined
         if "query" in locals() and query is not None and isinstance(query, dict):
@@ -82,6 +121,9 @@ async def process_query(
             print(f"{Colors.BROWN}WARN{Colors.END}: {Colors.RED}<-  {Colors.END}", end="")
             print(f"{Colors.RED}{e}{Colors.END}")
 
+        if raw_response:
+            raise e
+
         return templates.TemplateResponse(
             template,
             {
@@ -95,37 +137,6 @@ async def process_query(
             },
         )
 
-    if len(content) > MAX_DISPLAY_SIZE:
-        content = (
-            f"(Files content cropped to {int(MAX_DISPLAY_SIZE / 1_000)}k characters, "
-            "download full ingest to see more)\n" + content[:MAX_DISPLAY_SIZE]
-        )
-
-    _print_success(
-        url=query["url"],
-        max_file_size=max_file_size,
-        pattern_type=pattern_type,
-        pattern=pattern,
-        summary=summary,
-    )
-
-    return templates.TemplateResponse(
-        template,
-        {
-            "request": request,
-            "github_url": input_text,
-            "result": True,
-            "summary": summary,
-            "tree": tree,
-            "content": content,
-            "examples": EXAMPLE_REPOS if is_index else [],
-            "ingest_id": query["id"],
-            "default_file_size": slider_position,
-            "pattern_type": pattern_type,
-            "pattern": pattern,
-        },
-    )
-
 
 def _print_query(url: str, max_file_size: int, pattern_type: str, pattern: str) -> None:
     """

src/routers/dynamic.py

@@ -1,5 +1,7 @@
+import typing
+
 from fastapi import APIRouter, Form, Request
-from fastapi.responses import HTMLResponse
+from fastapi.responses import HTMLResponse, Response
 from fastapi.templating import Jinja2Templates
 
 from process_query import process_query
@@ -9,6 +11,45 @@
 templates = Jinja2Templates(directory="templates")
 
 
+@router.get("/extract/{full_path:path}", response_model=None)
+async def extract_content(
+    request: Request,
+    full_path: str,
+    summary: bool = False,
+) -> Response:
+    try:
+        query_params = request.query_params
+        max_file_size = int(query_params.get("max_file_size", 243))
+        pattern_type = query_params.get("pattern_type", "exclude")
+        pattern = query_params.get("pattern", "")
+
+        processed_query = await process_query(
+            request,
+            input_text=f"https://github.com/{full_path}",
+            slider_position=max_file_size,
+            pattern_type=pattern_type,
+            pattern=pattern,
+            is_index=False,
+            raw_response=True,
+        )
+
+        result_summary, tree, content = typing.cast(tuple[str, str, str], processed_query)
+
+        response_parts = []
+        if summary:
+            response_parts.append(f"Summary:\n{result_summary}\n")
+            response_parts.append(f"Tree:\n{tree}\n")
+        response_parts.append(f"Content:\n{content}")
+
+        return Response(content="\n".join(response_parts), media_type="text/plain")
+    except Exception as e:
+        return Response(
+            content=f"Error during extraction: {str(e)}",
+            media_type="text/plain",
+            status_code=500,
+        )
+
+
 @router.get("/{full_path:path}")
 async def catch_all(request: Request, full_path: str) -> HTMLResponse:
     """
@@ -49,7 +90,7 @@ async def process_catch_all(
     max_file_size: int = Form(...),
     pattern_type: str = Form(...),
     pattern: str = Form(...),
-) -> HTMLResponse:
+) -> HTMLResponse | tuple[str, str, str]:
     """
     Processes the form submission with user input for query parameters.
 

src/routers/index.py

@@ -47,7 +47,7 @@ async def index_post(
     max_file_size: int = Form(...),
     pattern_type: str = Form(...),
     pattern: str = Form(...),
-) -> HTMLResponse:
+) -> HTMLResponse | tuple[str, str, str]:
     """
     Processes the form submission with user input for query parameters.