Add path selection search, hyperlink-aware previews and better .docx/.pptx preview

Author: utkarsh0493

CLAUDE.md

@@ -0,0 +1,78 @@
+# CLAUDE.md
+
+This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
+
+## What this app does
+
+FSS Navigator App is a local Streamlit application for browsing computational research project folders. It renders a tree-style directory explorer on the left and a file preview panel on the right. All file access is local — no files leave the machine.
+
+## Running the app
+
+From any project folder you want to inspect:
+
+```bash
+python /path/to/FSS-Navigator-App/run_app.py
+```
+
+`run_app.py` resolves the app path relative to itself and launches `streamlit run Computations/Code/app/app.py`.
+
+The app uses `Path.cwd()` as the default project directory, so launch location matters.
+
+## Running tests
+
+Tests live in `Computations/Code/tests/` and are run with pytest from the repo root:
+
+```bash
+pytest Computations/Code/tests
+```
+
+To run a single test file:
+
+```bash
+pytest Computations/Code/tests/test_previewer.py
+```
+
+`conftest.py` adds `Computations/Code/` to `sys.path` so `core.*` imports resolve without package installation.
+
+## Environment setup
+
+The conda environment is named `fssnav`:
+
+```bash
+conda env create -f environment.yml
+conda activate fssnav
+```
+
+Or install via pip:
+
+```bash
+pip install -r requirements.txt
+```
+
+## Architecture
+
+```
+run_app.py                          ← entry point, launches streamlit
+Computations/Code/
+  app/app.py                        ← full Streamlit UI (single file)
+  core/
+    tree_builder.py                 ← list_directory(), search_directory()
+    previewer.py                    ← get_file_type() + per-format readers
+  tests/
+    conftest.py                     ← sys.path setup for core imports
+    test_tree_builder.py
+    test_previewer.py
+Data/demo_project/                  ← sample project for manual testing
+```
+
+**`app.py`** owns all Streamlit state (`session_state.project_path`, `session_state.selected_path`, `session_state.expanded_dirs`) and all UI rendering. It calls into `core/` for filesystem logic and file reading.
+
+**`tree_builder.py`** filters out dotfiles, caches, and tool-specific directories via `IGNORED_DIRECTORIES` and `IGNORED_FILES` sets. `list_directory()` returns folders before files, both sorted case-insensitively.
+
+**`previewer.py`** maps file extensions to a preview type string (`"text"`, `"table"`, `"image"`, `"pdf"`, `"notebook"`, `"docx"`, `"pptx"`, `"unsupported"`), then each reader function handles its type. PDF rendering uses PyMuPDF (`fitz`); notebooks use `nbconvert` HTMLExporter.
+
+## Key constraints
+
+- No external server calls — this is a privacy-first local tool.
+- `app.py` uses `st.rerun()` after every tree interaction; keep state mutations before `st.rerun()` calls, never after.
+- Adding a new file type requires changes in two places: `TEXT_EXTENSIONS`/`TABLE_EXTENSIONS`/etc. in `previewer.py` AND the preview dispatch block in `app.py`.

Computations/Code/app/app.py

@@ -11,21 +11,21 @@
 import sys
 from pathlib import Path
 import streamlit as st
-import streamlit.components.v1 as components
 
 CURRENT_FILE = Path(__file__).resolve()
 CODE_DIR = CURRENT_FILE.parents[1]
 sys.path.append(str(CODE_DIR))
 
-from core.tree_builder import list_directory  # noqa: E402
+from core.tree_builder import list_directory, search_directory  # noqa: E402
 from core.previewer import (  # noqa: E402
     get_file_type,
     read_text_file,
     read_table_file,
     render_notebook_html,
     render_pdf_pages,
-    read_docx_file,
-    read_pptx_file,
+    render_docx_as_html,
+    render_pptx_as_html,
+    extract_pdf_links,
 )
 
 st.set_page_config(
@@ -208,13 +208,54 @@
         background: #94a3b8;
     }
 
+    /* Input labels */
+    .stTextInput label,
+    div[data-testid="stTextInputRootElement"] label {
+        color: #374151 !important;
+        font-size: 0.85rem !important;
+    }
+
     </style>
     """,
     unsafe_allow_html=True,
 )
 
+st.markdown(
+    f"""
+    <div style="
+        margin-bottom: 0.6rem;
+        padding-left: 0.1rem;
+    ">
+        <h1 style="
+            margin-bottom:0rem;
+        ">
+            🧭 FSS Navigator
+        </h1>
+    </div>
+    """,
+    unsafe_allow_html=True,
+)
+
+default_project_path = Path.cwd().resolve()
+
+if "project_path" not in st.session_state:
+    st.session_state.project_path = str(default_project_path)
+
+project_path_input = st.text_input(
+    "Project directory to inspect",
+    value=st.session_state.project_path,
+    placeholder="/path/to/project",
+    label_visibility="collapsed",
+)
+
+project_path = Path(project_path_input).expanduser().resolve()
+
+if not project_path.exists() or not project_path.is_dir():
+    st.error("The provided path does not exist or is not a directory.")
+    st.stop()
+
+st.session_state.project_path = str(project_path)
 
-project_path = Path.cwd().resolve()
 
 if "selected_path" not in st.session_state:
     st.session_state.selected_path = None
@@ -275,35 +316,48 @@ def render_explorer(path: Path, level: int = 0, prefix: str = "") -> None:
                 st.rerun()
 
 
-st.markdown(
-    f"""
-    <div style="
-        margin-bottom: 0.6rem;
-        padding-left: 0.1rem;
-    ">
-        <h1 style="
-            margin-bottom:0rem;
-        ">
-            🧭 FSS Navigator
-        </h1>
-    </div>
-    """,
-    unsafe_allow_html=True,
-)
 
 left_col, right_col = st.columns([1.2, 3], gap="small")
 
 with left_col:
     left_panel = st.container(key="left_panel")
     with left_panel:
         st.markdown("### 📂 Directory structure")
-        render_explorer(project_path)
+
+        search_query = st.text_input(
+            "Search in selected directory",
+            placeholder="Search files or folders...",
+            key="directory_search",
+            label_visibility="collapsed",
+        )
+
+        if search_query:
+            search_results = search_directory(project_path, search_query)
+
+            st.caption(f"{len(search_results)} result(s) found")
+
+            for result in search_results[:100]:
+                icon = "📁" if result.is_dir() else "📄"
+                relative_result = result.relative_to(project_path)
+
+                if st.button(
+                    f"{icon} {relative_result}",
+                    key=f"search_{result}",
+                    use_container_width=True,
+                ):
+                    st.session_state.selected_path = str(result)
+                    st.rerun()
+
+            if len(search_results) > 100:
+                st.warning("Showing first 100 results only.")
+        else:
+            render_explorer(project_path)
 
 with right_col:
     right_panel = st.container(key="right_panel")
     with right_panel:
         st.markdown("### 👁️ Preview panel")
-    
+
         selected_path_value = st.session_state.selected_path
 
         if selected_path_value is None:
@@ -346,11 +400,11 @@ def render_explorer(path: Path, level: int = 0, prefix: str = "") -> None:
 
                     if suffix in ["md", "rmd"]:
                         content = read_text_file(str(selected_path), max_chars=None)
-                        st.markdown(content, unsafe_allow_html=False)
+                        st.markdown(content, unsafe_allow_html=True)
 
                     elif suffix == "html":
                         content = read_text_file(str(selected_path), max_chars=None)
-                        components.html(content, height=800, scrolling=True)
+                        st.html(content)
 
                     elif suffix in [
                         "py",
@@ -383,32 +437,44 @@ def render_explorer(path: Path, level: int = 0, prefix: str = "") -> None:
                         for page_number, image in pdf_pages:
                             st.write(f"Page {page_number}")
                             st.image(image, width="stretch")
+                        
+                        pdf_links = extract_pdf_links(str(selected_path))
+
+                        if pdf_links:
+                            st.write("### Links detected in PDF")
+
+                            for link in pdf_links:
+                                page_label = f"Page {link['page']}"
+
+                                if link["uri"]:
+                                    st.markdown(f"- {page_label}: [{link['uri']}]({link['uri']})")
+
+                                elif link["target_page"] is not None:
+                                    st.markdown(
+                                        f"- {page_label}: internal PDF link to page {link['target_page'] + 1}"
+                                    )
 
                     except Exception as error:
                         st.error(f"Could not render PDF: {error}")
 
                 elif preview_type == "notebook":
                     try:
                         notebook_html = render_notebook_html(str(selected_path))
-                        components.html(notebook_html, height=900, scrolling=True)
+                        st.html(notebook_html)
                     except Exception as error:
                         st.error(f"Could not render notebook: {error}")
 
                 elif preview_type == "docx":
                     try:
-                        docx_text = read_docx_file(str(selected_path))
-                        st.text_area(
-                            "DOCX content",
-                            value=docx_text,
-                            height=800,
-                        )
+                        docx_html = render_docx_as_html(str(selected_path))
+                        st.html(docx_html)
                     except Exception as error:
                         st.error(f"Could not preview DOCX file: {error}")
 
                 elif preview_type == "pptx":
                     try:
-                        pptx_text = read_pptx_file(str(selected_path))
-                        st.markdown(pptx_text)
+                        pptx_html = render_pptx_as_html(str(selected_path))
+                        st.html(pptx_html)
                     except Exception as error:
                         st.error(f"Could not preview PPTX file: {error}")