pdf-table-pull-lite 1.1.0__tar.gz

pdf_table_pull_lite-1.1.0/CHANGELOG.md

@@ -0,0 +1,15 @@
+# Changelog
+
+What changed in pdf-table-pull-lite, newest first.
+
+## 1.1.0 - 2026-11-04
+
+- New: a free MCP server, `pdf-table-pull-lite-mcp`. It gives AI apps one tool, `extract_pdf_tables`.
+- New: `llms.txt` with short notes for AI agents.
+- You can install it from PyPI with `pip install pdf-table-pull-lite`.
+
+## 1.0.0 - 2026-11-04
+
+- First release.
+- Pulls the tables out of a text PDF and saves each one as a CSV file.
+- Picks border or spacing mode for you, and tells you which pages are scans.

pdf_table_pull_lite-1.1.0/LICENSE

@@ -0,0 +1,9 @@
+MIT License
+
+Copyright (c) 2026 tidybyte
+
+Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

pdf_table_pull_lite-1.1.0/MANIFEST.in

@@ -0,0 +1,4 @@
+include CHANGELOG.md
+include llms.txt
+include server.json
+recursive-include tests *.py *.pdf *.html

pdf_table_pull_lite-1.1.0/PKG-INFO

@@ -0,0 +1,165 @@
+Metadata-Version: 2.4
+Name: pdf-table-pull-lite
+Version: 1.1.0
+Summary: Pull the tables out of a text PDF into CSV files. Runs locally. CLI and MCP server.
+Author: Tidy Byte Tools
+License-Expression: MIT
+Project-URL: Homepage, https://tidybyte.gumroad.com/l/hkxqxh
+Project-URL: Source, https://github.com/tidybytetools/pdf-table-pull-lite
+Project-URL: Issues, https://github.com/tidybytetools/pdf-table-pull-lite/issues
+Project-URL: Changelog, https://github.com/tidybytetools/pdf-table-pull-lite/blob/main/CHANGELOG.md
+Keywords: pdf,table,csv,extract,pdf-to-csv,pdfplumber,cli,mcp,mcp-server
+Classifier: Development Status :: 4 - Beta
+Classifier: Environment :: Console
+Classifier: Intended Audience :: Developers
+Classifier: Intended Audience :: End Users/Desktop
+Classifier: Operating System :: OS Independent
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3 :: Only
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Classifier: Topic :: Office/Business
+Classifier: Topic :: Text Processing
+Classifier: Topic :: Utilities
+Requires-Python: >=3.10
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: pdfplumber>=0.10
+Requires-Dist: mcp<2,>=1.2
+Dynamic: license-file
+
+# pdf-table-pull-lite
+
+[![License: MIT](https://img.shields.io/badge/license-MIT-green.svg)](https://github.com/tidybytetools/pdf-table-pull-lite/blob/main/LICENSE)
+[![Python 3.10+](https://img.shields.io/badge/python-3.10%2B-blue.svg)](https://www.python.org/)
+[![MCP server](https://img.shields.io/badge/MCP-stdio-purple.svg)](#use-it-from-claude-desktop-or-cursor-mcp)
+
+Free tool that pulls the tables out of a text PDF and saves each one as a CSV file.
+Need Excel output, scanned PDFs or tables joined across pages? The paid version does that: **[pdf-table-pull on Gumroad](https://tidybyte.gumroad.com/l/hkxqxh)**
+
+Runs on your own computer, so your PDFs are never uploaded anywhere. MIT license. Python 3.10 or newer, on Windows, Mac and Linux.
+
+<!-- mcp-name: io.github.tidybytetools/pdf-table-pull-lite -->
+
+## Install
+
+```
+pip install pdf-table-pull-lite
+```
+
+Or use pipx, which keeps it in its own space:
+
+```
+pipx install pdf-table-pull-lite
+```
+
+Want the newest code from GitHub instead?
+
+```
+pip install git+https://github.com/tidybytetools/pdf-table-pull-lite
+```
+
+## Use it
+
+```
+pdf-table-pull-lite statement.pdf                 # CSVs go in statement_tables/
+pdf-table-pull-lite statement.pdf -o out/ -p 2-5  # only pages 2 to 5
+pdf-table-pull-lite statement.pdf -s text         # tables with no border lines
+```
+
+Example run:
+
+```
+Found 2 table(s).
+  page1_table1: 2 rows x 3 columns
+  page1_table2: 2 rows x 2 columns
+Saved: statement_tables/page1_table1.csv
+Saved: statement_tables/page1_table2.csv
+```
+
+You get one CSV per table, named after its page. The CSVs open straight in Excel (they're saved as UTF-8 with a BOM so accents show correctly).
+
+`--strategy` (`-s`): `lines` is for tables with ruled borders, `text` is for tables lined up by spacing, and `auto` (the default) tries lines first, then text.
+
+When you run it in a terminal it ends with one short line about the paid version. `--quiet` turns it off, and it never shows when output is piped or redirected.
+
+## Known limits
+
+- **Scanned PDFs.** Pages that are just an image (no text layer) are skipped, and the tool tells you which ones.
+- **Tables that continue onto the next page** come out as separate CSVs, one per page.
+- **Odd layouts.** Tables with merged cells or no clear columns can come out with cells in the wrong spot. Check the CSV.
+- **CSV only.** No Excel or JSON output, no number cleanup (so `(40)` stays as text), and no password-protected PDFs.
+
+## Lite vs paid
+
+| | lite (free) | [pdf-table-pull](https://tidybyte.gumroad.com/l/hkxqxh) |
+|---|---|---|
+| Text PDFs, ruled or spaced tables | yes | yes |
+| CSV output | yes | yes |
+| Excel (one sheet per table) and JSON output | no | yes |
+| Scanned PDFs (OCR) | no | yes |
+| Join tables across pages | no | yes |
+| Numbers cleaned (`1,200` → 1200, `(40)` → -40) | no | yes |
+| Password-protected PDFs | no | yes |
+
+## How it compares to other free tools
+
+All of these are free and open source. Pick what fits you.
+
+| | pdf-table-pull-lite | [pdfplumber](https://github.com/jsvine/pdfplumber) | [Camelot](https://github.com/camelot-dev/camelot) | [tabula-py](https://github.com/chezou/tabula-py) |
+|---|---|---|---|---|
+| What it is | Command line tool | Python library | Python library + CLI | Python wrapper for Tabula |
+| Needs Java | no | no | no | yes |
+| One command to get CSVs | yes | no, you write code | yes | yes |
+| Picks border or spacing mode for you | yes (`auto`) | no | no | yes (guess) |
+| Tells you which pages are scans | yes | no | no | no |
+| Built-in MCP server for AI apps | yes | no | no | no |
+| Lots of tuning knobs | no | yes | yes | some |
+
+This tool is built on pdfplumber. If you want full control in your own code, use pdfplumber or Camelot directly.
+
+## Use it from Claude Desktop or Cursor (MCP)
+
+There is a small free MCP server built in. It gives your AI app one tool, `extract_pdf_tables`. It needs [uv](https://docs.astral.sh/uv/).
+
+Run it by hand to check it starts (it waits quietly for an app to talk to it):
+
+```
+uvx --from pdf-table-pull-lite pdf-table-pull-lite-mcp
+```
+
+Claude Desktop (`claude_desktop_config.json`) or Cursor (`~/.cursor/mcp.json`):
+
+```json
+{
+  "mcpServers": {
+    "pdf-table-pull-lite": {
+      "command": "uvx",
+      "args": ["--from", "pdf-table-pull-lite", "pdf-table-pull-lite-mcp"]
+    }
+  }
+}
+```
+
+Tool: `extract_pdf_tables(path, pages="", strategy="auto", output_dir="")`. It returns a list with the CSV text of each table. If you set `output_dir`, it saves the CSVs there and returns the file paths instead.
+
+## For AI agents
+
+- **Use when:** you have a local PDF with real text (you can select the words) and need its tables as rows.
+- **Don't use when:** the PDF is a scan or photo, it has a password, or you need Excel. Those pages get skipped. Use the paid version for those.
+- **Input:** a file path, plus optional `pages` like `1,3-5` and `strategy` (`auto`, `lines`, `text`).
+- **Output:** a list of CSV strings, one per table, first row is the header. Empty list means no tables were found.
+- CLI exit code is 0 on success and 1 on error (message on stderr). See also [`llms.txt`](https://github.com/tidybytetools/pdf-table-pull-lite/blob/main/llms.txt).
+
+## Tests
+
+```
+pip install pdfplumber "mcp<2"
+python3 -m unittest discover tests
+```
+
+## License
+
+MIT. See `LICENSE`.

pdf_table_pull_lite-1.1.0/README.md

@@ -0,0 +1,133 @@
+# pdf-table-pull-lite
+
+[![License: MIT](https://img.shields.io/badge/license-MIT-green.svg)](https://github.com/tidybytetools/pdf-table-pull-lite/blob/main/LICENSE)
+[![Python 3.10+](https://img.shields.io/badge/python-3.10%2B-blue.svg)](https://www.python.org/)
+[![MCP server](https://img.shields.io/badge/MCP-stdio-purple.svg)](#use-it-from-claude-desktop-or-cursor-mcp)
+
+Free tool that pulls the tables out of a text PDF and saves each one as a CSV file.
+Need Excel output, scanned PDFs or tables joined across pages? The paid version does that: **[pdf-table-pull on Gumroad](https://tidybyte.gumroad.com/l/hkxqxh)**
+
+Runs on your own computer, so your PDFs are never uploaded anywhere. MIT license. Python 3.10 or newer, on Windows, Mac and Linux.
+
+<!-- mcp-name: io.github.tidybytetools/pdf-table-pull-lite -->
+
+## Install
+
+```
+pip install pdf-table-pull-lite
+```
+
+Or use pipx, which keeps it in its own space:
+
+```
+pipx install pdf-table-pull-lite
+```
+
+Want the newest code from GitHub instead?
+
+```
+pip install git+https://github.com/tidybytetools/pdf-table-pull-lite
+```
+
+## Use it
+
+```
+pdf-table-pull-lite statement.pdf                 # CSVs go in statement_tables/
+pdf-table-pull-lite statement.pdf -o out/ -p 2-5  # only pages 2 to 5
+pdf-table-pull-lite statement.pdf -s text         # tables with no border lines
+```
+
+Example run:
+
+```
+Found 2 table(s).
+  page1_table1: 2 rows x 3 columns
+  page1_table2: 2 rows x 2 columns
+Saved: statement_tables/page1_table1.csv
+Saved: statement_tables/page1_table2.csv
+```
+
+You get one CSV per table, named after its page. The CSVs open straight in Excel (they're saved as UTF-8 with a BOM so accents show correctly).
+
+`--strategy` (`-s`): `lines` is for tables with ruled borders, `text` is for tables lined up by spacing, and `auto` (the default) tries lines first, then text.
+
+When you run it in a terminal it ends with one short line about the paid version. `--quiet` turns it off, and it never shows when output is piped or redirected.
+
+## Known limits
+
+- **Scanned PDFs.** Pages that are just an image (no text layer) are skipped, and the tool tells you which ones.
+- **Tables that continue onto the next page** come out as separate CSVs, one per page.
+- **Odd layouts.** Tables with merged cells or no clear columns can come out with cells in the wrong spot. Check the CSV.
+- **CSV only.** No Excel or JSON output, no number cleanup (so `(40)` stays as text), and no password-protected PDFs.
+
+## Lite vs paid
+
+| | lite (free) | [pdf-table-pull](https://tidybyte.gumroad.com/l/hkxqxh) |
+|---|---|---|
+| Text PDFs, ruled or spaced tables | yes | yes |
+| CSV output | yes | yes |
+| Excel (one sheet per table) and JSON output | no | yes |
+| Scanned PDFs (OCR) | no | yes |
+| Join tables across pages | no | yes |
+| Numbers cleaned (`1,200` → 1200, `(40)` → -40) | no | yes |
+| Password-protected PDFs | no | yes |
+
+## How it compares to other free tools
+
+All of these are free and open source. Pick what fits you.
+
+| | pdf-table-pull-lite | [pdfplumber](https://github.com/jsvine/pdfplumber) | [Camelot](https://github.com/camelot-dev/camelot) | [tabula-py](https://github.com/chezou/tabula-py) |
+|---|---|---|---|---|
+| What it is | Command line tool | Python library | Python library + CLI | Python wrapper for Tabula |
+| Needs Java | no | no | no | yes |
+| One command to get CSVs | yes | no, you write code | yes | yes |
+| Picks border or spacing mode for you | yes (`auto`) | no | no | yes (guess) |
+| Tells you which pages are scans | yes | no | no | no |
+| Built-in MCP server for AI apps | yes | no | no | no |
+| Lots of tuning knobs | no | yes | yes | some |
+
+This tool is built on pdfplumber. If you want full control in your own code, use pdfplumber or Camelot directly.
+
+## Use it from Claude Desktop or Cursor (MCP)
+
+There is a small free MCP server built in. It gives your AI app one tool, `extract_pdf_tables`. It needs [uv](https://docs.astral.sh/uv/).
+
+Run it by hand to check it starts (it waits quietly for an app to talk to it):
+
+```
+uvx --from pdf-table-pull-lite pdf-table-pull-lite-mcp
+```
+
+Claude Desktop (`claude_desktop_config.json`) or Cursor (`~/.cursor/mcp.json`):
+
+```json
+{
+  "mcpServers": {
+    "pdf-table-pull-lite": {
+      "command": "uvx",
+      "args": ["--from", "pdf-table-pull-lite", "pdf-table-pull-lite-mcp"]
+    }
+  }
+}
+```
+
+Tool: `extract_pdf_tables(path, pages="", strategy="auto", output_dir="")`. It returns a list with the CSV text of each table. If you set `output_dir`, it saves the CSVs there and returns the file paths instead.
+
+## For AI agents
+
+- **Use when:** you have a local PDF with real text (you can select the words) and need its tables as rows.
+- **Don't use when:** the PDF is a scan or photo, it has a password, or you need Excel. Those pages get skipped. Use the paid version for those.
+- **Input:** a file path, plus optional `pages` like `1,3-5` and `strategy` (`auto`, `lines`, `text`).
+- **Output:** a list of CSV strings, one per table, first row is the header. Empty list means no tables were found.
+- CLI exit code is 0 on success and 1 on error (message on stderr). See also [`llms.txt`](https://github.com/tidybytetools/pdf-table-pull-lite/blob/main/llms.txt).
+
+## Tests
+
+```
+pip install pdfplumber "mcp<2"
+python3 -m unittest discover tests
+```
+
+## License
+
+MIT. See `LICENSE`.

pdf_table_pull_lite-1.1.0/llms.txt

@@ -0,0 +1,18 @@
+# pdf-table-pull-lite
+
+> Free, MIT-licensed tool that pulls tables out of local text PDFs and returns them as CSV. Runs offline. Has a CLI, a Python API and a stdio MCP server.
+
+- Install: `pip install pdf-table-pull-lite` (or from GitHub: `pip install git+https://github.com/tidybytetools/pdf-table-pull-lite`)
+- CLI: `pdf-table-pull-lite file.pdf -o out/ -p 1-3 -s auto` writes one CSV per table (page<N>_table<K>.csv). Exit 0 ok, 1 error.
+- Python: `from pdf_table_pull_lite import extract; tables, scanned_pages = extract("file.pdf")` where each table is {"page", "index", "rows"}.
+- MCP: `uvx --from pdf-table-pull-lite pdf-table-pull-lite-mcp` exposes `extract_pdf_tables(path, pages="", strategy="auto", output_dir="") -> list[str]` (CSV text, or file paths if output_dir is set).
+
+## Use when
+- The PDF has a text layer and you need its tables as rows.
+
+## Don't use when
+- The PDF is scanned or image-only (pages are skipped), password-protected, or you need Excel/JSON or tables joined across pages. The paid version does those: https://tidybyte.gumroad.com/l/hkxqxh
+
+## Links
+- [README](https://github.com/tidybytetools/pdf-table-pull-lite/blob/main/README.md)
+- [Paid version](https://tidybyte.gumroad.com/l/hkxqxh)

pdf_table_pull_lite-1.1.0/pdf_table_pull_lite.egg-info/PKG-INFO

@@ -0,0 +1,165 @@
+Metadata-Version: 2.4
+Name: pdf-table-pull-lite
+Version: 1.1.0
+Summary: Pull the tables out of a text PDF into CSV files. Runs locally. CLI and MCP server.
+Author: Tidy Byte Tools
+License-Expression: MIT
+Project-URL: Homepage, https://tidybyte.gumroad.com/l/hkxqxh
+Project-URL: Source, https://github.com/tidybytetools/pdf-table-pull-lite
+Project-URL: Issues, https://github.com/tidybytetools/pdf-table-pull-lite/issues
+Project-URL: Changelog, https://github.com/tidybytetools/pdf-table-pull-lite/blob/main/CHANGELOG.md
+Keywords: pdf,table,csv,extract,pdf-to-csv,pdfplumber,cli,mcp,mcp-server
+Classifier: Development Status :: 4 - Beta
+Classifier: Environment :: Console
+Classifier: Intended Audience :: Developers
+Classifier: Intended Audience :: End Users/Desktop
+Classifier: Operating System :: OS Independent
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3 :: Only
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Classifier: Topic :: Office/Business
+Classifier: Topic :: Text Processing
+Classifier: Topic :: Utilities
+Requires-Python: >=3.10
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: pdfplumber>=0.10
+Requires-Dist: mcp<2,>=1.2
+Dynamic: license-file
+
+# pdf-table-pull-lite
+
+[![License: MIT](https://img.shields.io/badge/license-MIT-green.svg)](https://github.com/tidybytetools/pdf-table-pull-lite/blob/main/LICENSE)
+[![Python 3.10+](https://img.shields.io/badge/python-3.10%2B-blue.svg)](https://www.python.org/)
+[![MCP server](https://img.shields.io/badge/MCP-stdio-purple.svg)](#use-it-from-claude-desktop-or-cursor-mcp)
+
+Free tool that pulls the tables out of a text PDF and saves each one as a CSV file.
+Need Excel output, scanned PDFs or tables joined across pages? The paid version does that: **[pdf-table-pull on Gumroad](https://tidybyte.gumroad.com/l/hkxqxh)**
+
+Runs on your own computer, so your PDFs are never uploaded anywhere. MIT license. Python 3.10 or newer, on Windows, Mac and Linux.
+
+<!-- mcp-name: io.github.tidybytetools/pdf-table-pull-lite -->
+
+## Install
+
+```
+pip install pdf-table-pull-lite
+```
+
+Or use pipx, which keeps it in its own space:
+
+```
+pipx install pdf-table-pull-lite
+```
+
+Want the newest code from GitHub instead?
+
+```
+pip install git+https://github.com/tidybytetools/pdf-table-pull-lite
+```
+
+## Use it
+
+```
+pdf-table-pull-lite statement.pdf                 # CSVs go in statement_tables/
+pdf-table-pull-lite statement.pdf -o out/ -p 2-5  # only pages 2 to 5
+pdf-table-pull-lite statement.pdf -s text         # tables with no border lines
+```
+
+Example run:
+
+```
+Found 2 table(s).
+  page1_table1: 2 rows x 3 columns
+  page1_table2: 2 rows x 2 columns
+Saved: statement_tables/page1_table1.csv
+Saved: statement_tables/page1_table2.csv
+```
+
+You get one CSV per table, named after its page. The CSVs open straight in Excel (they're saved as UTF-8 with a BOM so accents show correctly).
+
+`--strategy` (`-s`): `lines` is for tables with ruled borders, `text` is for tables lined up by spacing, and `auto` (the default) tries lines first, then text.
+
+When you run it in a terminal it ends with one short line about the paid version. `--quiet` turns it off, and it never shows when output is piped or redirected.
+
+## Known limits
+
+- **Scanned PDFs.** Pages that are just an image (no text layer) are skipped, and the tool tells you which ones.
+- **Tables that continue onto the next page** come out as separate CSVs, one per page.
+- **Odd layouts.** Tables with merged cells or no clear columns can come out with cells in the wrong spot. Check the CSV.
+- **CSV only.** No Excel or JSON output, no number cleanup (so `(40)` stays as text), and no password-protected PDFs.
+
+## Lite vs paid
+
+| | lite (free) | [pdf-table-pull](https://tidybyte.gumroad.com/l/hkxqxh) |
+|---|---|---|
+| Text PDFs, ruled or spaced tables | yes | yes |
+| CSV output | yes | yes |
+| Excel (one sheet per table) and JSON output | no | yes |
+| Scanned PDFs (OCR) | no | yes |
+| Join tables across pages | no | yes |
+| Numbers cleaned (`1,200` → 1200, `(40)` → -40) | no | yes |
+| Password-protected PDFs | no | yes |
+
+## How it compares to other free tools
+
+All of these are free and open source. Pick what fits you.
+
+| | pdf-table-pull-lite | [pdfplumber](https://github.com/jsvine/pdfplumber) | [Camelot](https://github.com/camelot-dev/camelot) | [tabula-py](https://github.com/chezou/tabula-py) |
+|---|---|---|---|---|
+| What it is | Command line tool | Python library | Python library + CLI | Python wrapper for Tabula |
+| Needs Java | no | no | no | yes |
+| One command to get CSVs | yes | no, you write code | yes | yes |
+| Picks border or spacing mode for you | yes (`auto`) | no | no | yes (guess) |
+| Tells you which pages are scans | yes | no | no | no |
+| Built-in MCP server for AI apps | yes | no | no | no |
+| Lots of tuning knobs | no | yes | yes | some |
+
+This tool is built on pdfplumber. If you want full control in your own code, use pdfplumber or Camelot directly.
+
+## Use it from Claude Desktop or Cursor (MCP)
+
+There is a small free MCP server built in. It gives your AI app one tool, `extract_pdf_tables`. It needs [uv](https://docs.astral.sh/uv/).
+
+Run it by hand to check it starts (it waits quietly for an app to talk to it):
+
+```
+uvx --from pdf-table-pull-lite pdf-table-pull-lite-mcp
+```
+
+Claude Desktop (`claude_desktop_config.json`) or Cursor (`~/.cursor/mcp.json`):
+
+```json
+{
+  "mcpServers": {
+    "pdf-table-pull-lite": {
+      "command": "uvx",
+      "args": ["--from", "pdf-table-pull-lite", "pdf-table-pull-lite-mcp"]
+    }
+  }
+}
+```
+
+Tool: `extract_pdf_tables(path, pages="", strategy="auto", output_dir="")`. It returns a list with the CSV text of each table. If you set `output_dir`, it saves the CSVs there and returns the file paths instead.
+
+## For AI agents
+
+- **Use when:** you have a local PDF with real text (you can select the words) and need its tables as rows.
+- **Don't use when:** the PDF is a scan or photo, it has a password, or you need Excel. Those pages get skipped. Use the paid version for those.
+- **Input:** a file path, plus optional `pages` like `1,3-5` and `strategy` (`auto`, `lines`, `text`).
+- **Output:** a list of CSV strings, one per table, first row is the header. Empty list means no tables were found.
+- CLI exit code is 0 on success and 1 on error (message on stderr). See also [`llms.txt`](https://github.com/tidybytetools/pdf-table-pull-lite/blob/main/llms.txt).
+
+## Tests
+
+```
+pip install pdfplumber "mcp<2"
+python3 -m unittest discover tests
+```
+
+## License
+
+MIT. See `LICENSE`.

pdf_table_pull_lite-1.1.0/pdf_table_pull_lite.egg-info/SOURCES.txt

@@ -0,0 +1,21 @@
+CHANGELOG.md
+LICENSE
+MANIFEST.in
+README.md
+llms.txt
+pdf_table_pull_lite.py
+pdf_table_pull_lite_mcp.py
+pyproject.toml
+server.json
+pdf_table_pull_lite.egg-info/PKG-INFO
+pdf_table_pull_lite.egg-info/SOURCES.txt
+pdf_table_pull_lite.egg-info/dependency_links.txt
+pdf_table_pull_lite.egg-info/entry_points.txt
+pdf_table_pull_lite.egg-info/requires.txt
+pdf_table_pull_lite.egg-info/top_level.txt
+tests/test_lite.py
+tests/test_mcp.py
+tests/fixtures/empty.pdf
+tests/fixtures/plain.pdf
+tests/fixtures/ruled.pdf
+tests/fixtures/scanned.pdf

pdf_table_pull_lite-1.1.0/pdf_table_pull_lite.egg-info/dependency_links.txt

@@ -0,0 +1 @@
+

pdf_table_pull_lite-1.1.0/pdf_table_pull_lite.egg-info/entry_points.txt

@@ -0,0 +1,3 @@
+[console_scripts]
+pdf-table-pull-lite = pdf_table_pull_lite:main
+pdf-table-pull-lite-mcp = pdf_table_pull_lite_mcp:main

pdf_table_pull_lite-1.1.0/pdf_table_pull_lite.egg-info/requires.txt

@@ -0,0 +1,2 @@
+pdfplumber>=0.10
+mcp<2,>=1.2

pdf_table_pull_lite-1.1.0/pdf_table_pull_lite.egg-info/top_level.txt

@@ -0,0 +1,2 @@
+pdf_table_pull_lite
+pdf_table_pull_lite_mcp

pdf_table_pull_lite-1.1.0/pdf_table_pull_lite.py

@@ -0,0 +1,131 @@
+#!/usr/bin/env python3
+"""pdf-table-pull-lite: pull the tables out of a text PDF into CSV files.
+
+Free, MIT-licensed lite edition. The paid version (pdf-table-pull) adds Excel and
+JSON output, scanned-PDF OCR, joining tables across pages, and number cleanup:
+https://tidybyte.gumroad.com/l/hkxqxh
+
+Usage:
+  pdf-table-pull-lite report.pdf
+  pdf-table-pull-lite report.pdf -o out_dir/ --pages 2-5
+"""
+import argparse, csv, os, re, sys
+
+__version__ = "1.1.0"
+PRO_URL = "https://tidybyte.gumroad.com/l/hkxqxh"
+PRO_LINE = f"Need Excel output, scanned PDFs or tables joined across pages? pdf-table-pull (paid): {PRO_URL}"
+
+
+def parse_pages(spec, total):
+    """'1,3-5' -> [1,3,4,5] (1-based). None means every page."""
+    if not spec:
+        return list(range(1, total + 1))
+    pages = []
+    for part in spec.split(","):
+        part = part.strip()
+        if not part:
+            continue
+        if "-" in part:
+            a, b = part.split("-", 1)
+            pages.extend(range(int(a) if a else 1, (int(b) if b else total) + 1))
+        else:
+            pages.append(int(part))
+    bad = [p for p in pages if p < 1 or p > total]
+    if bad:
+        raise ValueError(f"page {bad[0]} is out of range (PDF has {total} pages)")
+    return sorted(set(pages))
+
+
+def _clean_cell(c):
+    return "" if c is None else re.sub(r"\s+", " ", str(c)).strip()
+
+
+def _clean_table(rows):
+    rows = [[_clean_cell(c) for c in r] for r in rows]
+    rows = [r for r in rows if any(r)]
+    if not rows:
+        return []
+    width = max(len(r) for r in rows)
+    rows = [r + [""] * (width - len(r)) for r in rows]
+    keep = [i for i in range(width) if any(r[i] for r in rows)]
+    return [[r[i] for i in keep] for r in rows]
+
+
+def extract(pdf_path, pages=None, strategy="auto"):
+    """Return (tables, scanned_pages). tables: list of {'page', 'index', 'rows'}.
+    Pages with no text layer (scans) are skipped and listed in scanned_pages."""
+    import pdfplumber
+    if strategy not in ("auto", "lines", "text"):
+        raise ValueError("strategy must be lines, text or auto")
+    lines = {"vertical_strategy": "lines", "horizontal_strategy": "lines"}
+    text = {"vertical_strategy": "text", "horizontal_strategy": "text"}
+    out, scanned = [], []
+    with pdfplumber.open(pdf_path) as pdf:
+        for pno in parse_pages(pages, len(pdf.pages)):
+            page = pdf.pages[pno - 1]
+            if not page.chars:
+                scanned.append(pno)
+                continue
+            if strategy == "auto":
+                found = page.extract_tables(lines)
+                if not any(_clean_table(t) for t in found):
+                    found = page.extract_tables(text)
+            else:
+                found = page.extract_tables(lines if strategy == "lines" else text)
+            k = 0
+            for t in found:
+                rows = _clean_table(t)
+                if not rows or (len(rows) < 2 and strategy != "lines"):
+                    continue
+                k += 1
+                out.append({"page": pno, "index": k, "rows": rows})
+    return out, scanned
+
+
+def write_csv(tables, out_dir):
+    os.makedirs(out_dir, exist_ok=True)
+    paths = []
+    for t in tables:
+        p = os.path.join(out_dir, f"page{t['page']}_table{t['index']}.csv")
+        with open(p, "w", newline="", encoding="utf-8-sig") as f:
+            csv.writer(f).writerows(t["rows"])
+        paths.append(p)
+    return paths
+
+
+def main(argv=None):
+    ap = argparse.ArgumentParser(prog="pdf-table-pull-lite",
+                                 description="Pull tables out of a text PDF into CSV files (one file per table).",
+                                 epilog=PRO_LINE)
+    ap.add_argument("pdf")
+    ap.add_argument("-o", "--output", help="output folder (default: <pdf name>_tables next to the PDF)")
+    ap.add_argument("-p", "--pages", help="pages to read, e.g. 1,3-5 (default: all)")
+    ap.add_argument("-s", "--strategy", choices=["auto", "lines", "text"], default="auto",
+                    help="lines = tables with ruled borders, text = tables lined up by spacing, auto = try lines then text")
+    ap.add_argument("-q", "--quiet", action="store_true", help="don't print the one-line note about the paid version")
+    ap.add_argument("--version", action="version", version=__version__)
+    a = ap.parse_args(argv)
+    try:
+        if not os.path.isfile(a.pdf):
+            raise FileNotFoundError(f"file not found: {a.pdf}")
+        tables, scanned = extract(a.pdf, a.pages, a.strategy)
+        files = write_csv(tables, a.output or os.path.splitext(a.pdf)[0] + "_tables")
+        print(f"Found {len(tables)} table(s).")
+        for t in tables:
+            print(f"  page{t['page']}_table{t['index']}: {len(t['rows']) - 1} rows x {len(t['rows'][0])} columns")
+        for p in files:
+            print(f"Saved: {p}")
+        if scanned:
+            print(f"Skipped page(s) {', '.join(map(str, scanned))}: no text layer (scanned image). The lite version reads text PDFs only.")
+        elif not tables:
+            print("Tip: no tables found. Try --strategy text.")
+        if not a.quiet and sys.stdout.isatty():
+            print(PRO_LINE)
+        return 0
+    except Exception as e:
+        print(f"Error: {e}", file=sys.stderr)
+        return 1
+
+
+if __name__ == "__main__":
+    sys.exit(main())

pdf_table_pull_lite-1.1.0/pdf_table_pull_lite_mcp.py

@@ -0,0 +1,43 @@
+"""Free MCP server for pdf-table-pull-lite (stdio).
+
+Run:  uvx --from git+https://github.com/tidybytetools/pdf-table-pull-lite pdf-table-pull-lite-mcp
+"""
+import csv, io, os
+from mcp.server.fastmcp import FastMCP
+import pdf_table_pull_lite as lite
+
+mcp = FastMCP("pdf-table-pull-lite")
+
+
+@mcp.tool()
+def extract_pdf_tables(path: str, pages: str = "", strategy: str = "auto", output_dir: str = "") -> list[str]:
+    """Pull the tables out of a local text PDF.
+
+    path: full path to a PDF on this computer.
+    pages: optional, like "1,3-5". Empty means all pages.
+    strategy: "auto" (default), "lines" (ruled borders) or "text" (lined up by spacing).
+    output_dir: optional. If set, each table is saved there as a CSV and the file paths
+    are returned. If empty, the CSV text of each table is returned.
+
+    Scanned (image-only) pages have no text and are skipped. Returns [] if no tables.
+    """
+    path = os.path.expanduser(path)
+    if not os.path.isfile(path):
+        raise FileNotFoundError(f"file not found: {path}")
+    tables, _scanned = lite.extract(path, pages or None, strategy)
+    if output_dir:
+        return lite.write_csv(tables, os.path.expanduser(output_dir))
+    out = []
+    for t in tables:
+        buf = io.StringIO()
+        csv.writer(buf).writerows(t["rows"])
+        out.append(buf.getvalue())
+    return out
+
+
+def main():
+    mcp.run()
+
+
+if __name__ == "__main__":
+    main()

pdf_table_pull_lite-1.1.0/pyproject.toml

@@ -0,0 +1,47 @@
+[build-system]
+requires = ["setuptools>=77"]
+build-backend = "setuptools.build_meta"
+
+[project]
+name = "pdf-table-pull-lite"
+version = "1.1.0"
+description = "Pull the tables out of a text PDF into CSV files. Runs locally. CLI and MCP server."
+readme = "README.md"
+license = "MIT"
+license-files = ["LICENSE"]
+authors = [{name = "Tidy Byte Tools"}]
+requires-python = ">=3.10"
+dependencies = [
+  "pdfplumber>=0.10",
+  "mcp>=1.2,<2",
+]
+keywords = ["pdf", "table", "csv", "extract", "pdf-to-csv", "pdfplumber", "cli", "mcp", "mcp-server"]
+classifiers = [
+  "Development Status :: 4 - Beta",
+  "Environment :: Console",
+  "Intended Audience :: Developers",
+  "Intended Audience :: End Users/Desktop",
+  "Operating System :: OS Independent",
+  "Programming Language :: Python :: 3",
+  "Programming Language :: Python :: 3 :: Only",
+  "Programming Language :: Python :: 3.10",
+  "Programming Language :: Python :: 3.11",
+  "Programming Language :: Python :: 3.12",
+  "Programming Language :: Python :: 3.13",
+  "Topic :: Office/Business",
+  "Topic :: Text Processing",
+  "Topic :: Utilities",
+]
+
+[project.urls]
+Homepage = "https://tidybyte.gumroad.com/l/hkxqxh"
+Source = "https://github.com/tidybytetools/pdf-table-pull-lite"
+Issues = "https://github.com/tidybytetools/pdf-table-pull-lite/issues"
+Changelog = "https://github.com/tidybytetools/pdf-table-pull-lite/blob/main/CHANGELOG.md"
+
+[project.scripts]
+pdf-table-pull-lite = "pdf_table_pull_lite:main"
+pdf-table-pull-lite-mcp = "pdf_table_pull_lite_mcp:main"
+
+[tool.setuptools]
+py-modules = ["pdf_table_pull_lite", "pdf_table_pull_lite_mcp"]

pdf_table_pull_lite-1.1.0/server.json

@@ -0,0 +1,24 @@
+{
+  "$schema": "https://static.modelcontextprotocol.io/schemas/2025-12-11/server.schema.json",
+  "name": "io.github.tidybytetools/pdf-table-pull-lite",
+  "title": "pdf-table-pull-lite",
+  "description": "Pull tables out of local text PDFs as CSV. Runs on your computer, nothing uploaded.",
+  "repository": {
+    "url": "https://github.com/tidybytetools/pdf-table-pull-lite",
+    "source": "github"
+  },
+  "websiteUrl": "https://github.com/tidybytetools/pdf-table-pull-lite",
+  "version": "1.1.0",
+  "packages": [
+    {
+      "registryType": "pypi",
+      "registryBaseUrl": "https://pypi.org",
+      "identifier": "pdf-table-pull-lite",
+      "version": "1.1.0",
+      "runtimeHint": "uvx",
+      "transport": {
+        "type": "stdio"
+      }
+    }
+  ]
+}

pdf_table_pull_lite-1.1.0/setup.cfg

@@ -0,0 +1,4 @@
+[egg_info]
+tag_build = 
+tag_date = 0
+

pdf_table_pull_lite-1.1.0/tests/fixtures/empty.pdf

@@ -0,0 +1,68 @@
+%PDF-1.3
+%���� ReportLab Generated PDF document (opensource)
+1 0 obj
+<<
+/F1 2 0 R
+>>
+endobj
+2 0 obj
+<<
+/BaseFont /Helvetica /Encoding /WinAnsiEncoding /Name /F1 /Subtype /Type1 /Type /Font
+>>
+endobj
+3 0 obj
+<<
+/Contents 7 0 R /MediaBox [ 0 0 612 792 ] /Parent 6 0 R /Resources <<
+/Font 1 0 R /ProcSet [ /PDF /Text /ImageB /ImageC /ImageI ]
+>> /Rotate 0 /Trans <<
+
+>> 
+  /Type /Page
+>>
+endobj
+4 0 obj
+<<
+/PageMode /UseNone /Pages 6 0 R /Type /Catalog
+>>
+endobj
+5 0 obj
+<<
+/Author (anonymous) /CreationDate (D:20261104013809-06'00') /Creator (anonymous) /Keywords () /ModDate (D:20261104013809-06'00') /Producer (ReportLab PDF Library - \(opensource\)) 
+  /Subject (unspecified) /Title (untitled) /Trapped /False
+>>
+endobj
+6 0 obj
+<<
+/Count 1 /Kids [ 3 0 R ] /Type /Pages
+>>
+endobj
+7 0 obj
+<<
+/Filter [ /ASCII85Decode /FlateDecode ] /Length 127
+>>
+stream
+Gap@D5mkI_'F"4o`@S&\bn#NpKt7h^rrOp&.:OKgo`!E>*mA)\-b+<)0kIjCkGo^HI3R]*U3@sO<\Je<4Ma$7e)-JFLf>nL9!T,l6]7#71#(<7"8Y?Fs$,<MnRNNY~>endstream
+endobj
+xref
+0 8
+0000000000 65535 f 
+0000000061 00000 n 
+0000000092 00000 n 
+0000000199 00000 n 
+0000000392 00000 n 
+0000000460 00000 n 
+0000000721 00000 n 
+0000000780 00000 n 
+trailer
+<<
+/ID 
+[<a4b14a0b3c7a153e35d150475c5cef87><a4b14a0b3c7a153e35d150475c5cef87>]
+% ReportLab generated PDF document -- digest (opensource)
+
+/Info 5 0 R
+/Root 4 0 R
+/Size 8
+>>
+startxref
+997
+%%EOF

pdf_table_pull_lite-1.1.0/tests/fixtures/plain.pdf

@@ -0,0 +1,68 @@
+%PDF-1.3
+%���� ReportLab Generated PDF document (opensource)
+1 0 obj
+<<
+/F1 2 0 R
+>>
+endobj
+2 0 obj
+<<
+/BaseFont /Helvetica /Encoding /WinAnsiEncoding /Name /F1 /Subtype /Type1 /Type /Font
+>>
+endobj
+3 0 obj
+<<
+/Contents 7 0 R /MediaBox [ 0 0 612 792 ] /Parent 6 0 R /Resources <<
+/Font 1 0 R /ProcSet [ /PDF /Text /ImageB /ImageC /ImageI ]
+>> /Rotate 0 /Trans <<
+
+>> 
+  /Type /Page
+>>
+endobj
+4 0 obj
+<<
+/PageMode /UseNone /Pages 6 0 R /Type /Catalog
+>>
+endobj
+5 0 obj
+<<
+/Author (anonymous) /CreationDate (D:20261104013809-06'00') /Creator (anonymous) /Keywords () /ModDate (D:20261104013809-06'00') /Producer (ReportLab PDF Library - \(opensource\)) 
+  /Subject (unspecified) /Title (untitled) /Trapped /False
+>>
+endobj
+6 0 obj
+<<
+/Count 1 /Kids [ 3 0 R ] /Type /Pages
+>>
+endobj
+7 0 obj
+<<
+/Filter [ /ASCII85Decode /FlateDecode ] /Length 220
+>>
+stream
+Gas3._$\%5%"R<K:Q_CZX%?@(GuUYcn7H!7Lk;p8E:s/]UJiPK"EFbir.G(K:qodX^d__:,9R3$!,88q%^;4T^Jgd/6\,Nk_R-?!Gm89kRK"CAI&j$KNsNqP%hn9ok!/H/;f]&YfjmVkE]cJ:os8_kHb;I#@T80R[*doUSk')jeP/tdM0*^;;Gj0ic+'hgj^@i6Ir/+k#=pmW`U_3gj9<h-E'O~>endstream
+endobj
+xref
+0 8
+0000000000 65535 f 
+0000000061 00000 n 
+0000000092 00000 n 
+0000000199 00000 n 
+0000000392 00000 n 
+0000000460 00000 n 
+0000000721 00000 n 
+0000000780 00000 n 
+trailer
+<<
+/ID 
+[<450000f58db8cd681e2fa7725f3a366d><450000f58db8cd681e2fa7725f3a366d>]
+% ReportLab generated PDF document -- digest (opensource)
+
+/Info 5 0 R
+/Root 4 0 R
+/Size 8
+>>
+startxref
+1090
+%%EOF

pdf_table_pull_lite-1.1.0/tests/fixtures/ruled.pdf

@@ -0,0 +1,112 @@
+%PDF-1.4
+%���� ReportLab Generated PDF document (opensource)
+1 0 obj
+<<
+/F1 2 0 R /F2 3 0 R
+>>
+endobj
+2 0 obj
+<<
+/BaseFont /Helvetica /Encoding /WinAnsiEncoding /Name /F1 /Subtype /Type1 /Type /Font
+>>
+endobj
+3 0 obj
+<<
+/BaseFont /Helvetica-Bold /Encoding /WinAnsiEncoding /Name /F2 /Subtype /Type1 /Type /Font
+>>
+endobj
+4 0 obj
+<<
+/Contents 10 0 R /MediaBox [ 0 0 612 792 ] /Parent 9 0 R /Resources <<
+/Font 1 0 R /ProcSet [ /PDF /Text /ImageB /ImageC /ImageI ]
+>> /Rotate 0 /Trans <<
+
+>> 
+  /Type /Page
+>>
+endobj
+5 0 obj
+<<
+/Contents 11 0 R /MediaBox [ 0 0 612 792 ] /Parent 9 0 R /Resources <<
+/Font 1 0 R /ProcSet [ /PDF /Text /ImageB /ImageC /ImageI ]
+>> /Rotate 0 /Trans <<
+
+>> 
+  /Type /Page
+>>
+endobj
+6 0 obj
+<<
+/Contents 12 0 R /MediaBox [ 0 0 612 792 ] /Parent 9 0 R /Resources <<
+/Font 1 0 R /ProcSet [ /PDF /Text /ImageB /ImageC /ImageI ]
+>> /Rotate 0 /Trans <<
+
+>> 
+  /Type /Page
+>>
+endobj
+7 0 obj
+<<
+/PageMode /UseNone /Pages 9 0 R /Type /Catalog
+>>
+endobj
+8 0 obj
+<<
+/Author (\(anonymous\)) /CreationDate (D:20261104013809-06'00') /Creator (\(unspecified\)) /Keywords () /ModDate (D:20261104013809-06'00') /Producer (ReportLab PDF Library - \(opensource\)) 
+  /Subject (\(unspecified\)) /Title (\(anonymous\)) /Trapped /False
+>>
+endobj
+9 0 obj
+<<
+/Count 3 /Kids [ 4 0 R 5 0 R 6 0 R ] /Type /Pages
+>>
+endobj
+10 0 obj
+<<
+/Filter [ /ASCII85Decode /FlateDecode ] /Length 491
+>>
+stream
+Gat=fd8,BK&B<W'/+.C5>'m(tj`QCuNEO-<j'I](F;[H.`U(A6g&`3$Yf%*=bK$gD?$@@lV[)&HCCW$kiV";n"f1U;P"A3:i@YSR"0BVcJ;sfLUiE;tl[Y"(lU5B+EY3iA\d2_-:rQ+sO?TR\0M1skAX/P8"-hJ?;-$Ud2iq@^9a@H-#sH>c6:MA0kcgu")k4IE76U:b<#R%9YpPBFrE8KUIU,La9Uc5[YTGWh74fb\)%@d3^msid1"NGgn60cpd]^VV21TC;%TsSEnW"ZKbud'RX<s+6VC.a:`"ThRqN"uG"AHXMYbPAna8n#)Og+KoQ;6R4Q$5WFUtY9c#,+6'<&ZR$X-K[XZ4-+(XcKYh)oLo<<d-ta-YMD4?n3>c:?3V@PZ4;BoC'md<`qVj'$b0u9q&*"KfI>7D=o^I<-)e&9MJ.Aqf*]R7@gYggqFOT`q/;=EH(ZskFU.?o3ORe]N<k$LXOIa9mkJo0EckRPm7m~>endstream
+endobj
+11 0 obj
+<<
+/Filter [ /ASCII85Decode /FlateDecode ] /Length 1081
+>>
+stream
+GasJT>Dpur'SPB,=7i9a0_QeaV]WUILnNl@LoeJQm,GU7J)?dg\Ig_G_^okL.ePA\V+;qe\A$Dbp\Q7iQV2T&2MIC5nfssh4G;g+GdEohT+WYqREf:Mo<@N3_f7bcp[]YC`pM!okH.89`$nBZL?0&>]RTl:/D"<U[Su71?;D9cVf.p&a)u5_(NXqC\UCuuIp>!fZ*d;k&5kp"*rasP?EW_5J\$Ka17lO2aIUd,b)jdjjW:@g/!c8YPtb4#QUT%[:bu0(+a4b^d/^^.;XT+)\RGlU\Yihg>S\d_5g<Q=K\Y3I0L9U3jm#+U>d6]PU%Jsr`"eDT&(I/NQB5<t#\M/[]3HATR4FSb48'4Ba;oK83`%V6/0VK%^;qJ-m"0@*B<msmFQ.b#SL#aQ_pU;=LlOV*$.bE@$k7R,9G+k[4Uh]UQ`(<%X340[HKp&PYe3A8B\6X'$IJr'!qqtZ$k3Tg9HcH<,%-+]b0[B1bS6cb-N^`DW!j,NbY/U02bjK,1/So-2OTt"FQGp+/eaMD@@<N47$!UO*?][``U$.@BJkM-dM%aORu,;7Yg(B+U,1[\_);slTjaFY.7D4JL^`$3</--EAu5j)k+7A;gItS$L2CP"&3Y"W@%3"qd'1d='VYh-"Dfkk)&frOR6l#S7)1\oQ(_WUR41NO:'>Ah;&L:\R@,7@j$:C8C\s40/lPE]Xd'tt&@_'?G,4P2[=gL:4UU:8&!B$BIi]lg'&O.A&\]LEAa;l=pO[4<Hb52TJaK'SEIWfkr1:M)5=b.?pb@p@G>S-)j2:`-k+=?A]K;\)nFq%2I\'.?Qh:u]cZ@rt?fC=ZTYo.t@0oc264QcVs)71FON#iV>L9%$3k<3jk[L(<@NBPF3&D)Ua::st/u"+kK+G;@e"DW@_.h98khn;]U,Y+A\k!<+K.KW9E#Ngq*Lm=>9FagM/d;QY5o84jC!]<@n7_;R4#d\\W")RT/k-)D5o7MVC!]:hE#N8LRY0(^MqPp\d)G)(2!ACfE#Nh\*Lm=>6k>=_DiWW^>ogJeYkbB20AM'7%dKI6b_:@nW;VocFtM03:Rh>T2P7H[3;ir2VKPZ~>endstream
+endobj
+12 0 obj
+<<
+/Filter [ /ASCII85Decode /FlateDecode ] /Length 932
+>>
+stream
+Gas2Kh+e\N&;9Mu;p;(aZ-Mh+cDCQg(__%!J/C*d*f(*lp"e<cl#$^&66N2(C[-PKW#V>4qhI8Ir\tkgH8@5346+!,IMs$#BXZ7Lh0/X9Ehr>9Sl.K'YPX%Pg`GXf^J=0r4g'+B\=^4!Z\_=/gY"^dq"D8OY0?U6;trKB(Rs^5`VK(=?!aEODQh>EpXaUb.sFucLFORj,h8C]^NVX"b?Hcc':MY_PjAr%QE$AsRT+k[a9g+g&FFL</J(W+5nE%0AlFMoY,YNY-C(5/E^I)]6AMrq,%Q@E.R=U+'eQ9o4kE*?JV)6i(UK6octj':YpLYULlPcb7WLi%.7F'%M!-#9G;1&Y$-f=Ym:mhSJ8!QRCa:AbRj,5c2%_PeFAiKl+KoH5L$44'!Sa]T":a*Z9HcGo,%-ltQ(VQTM((gt8d&t:V@k_%1GOSs:PjU=el76UD'l+kD(-($c'h8OYg(*#U"rlc0iCtm;_`SDY@oYEkd`#a_)?'Y6,PSo;i$Gk&IGmK;[FANR@$<_W/4'Lc'h8OYg(*#U0UJ]Jo0(<Jqe+hV?i_A&?sB!+#;J[Cl,Y0V#`L-dO,G=)-\;&U+!eq97=H\Mm4RW2crND*Ef!dr._:Uingg^>[VuaE.9`<pm_l>%*P\EeOWA0R1)/jq60#2rcps4"b$8eRZdJj]&$f7FZ/F]:N2k*Igj*:pUZaDI:44ffm9rBl:JU280<-(JQ4eGC%\auTT@Sd*5D;Z8["PFrUG4uG4\u<'M7>K+pG.cqnNRUeRBWS,%UABb.rp(IFDcc`N8!/i'gW71@W.2GSta72;APoY6nS??7ZFi#$rPY0*<kZJsV$):;P-#(+q+^)0iD9GSt`\.sMqV&J_qe]^Q87#$tC@[fXm5SHZ\Tnfm"f\4Ji&rO`kE1EXWc8iSi@23:k58Xlp(LH5;<%"X\]~>endstream
+endobj
+xref
+0 13
+0000000000 65535 f 
+0000000061 00000 n 
+0000000102 00000 n 
+0000000209 00000 n 
+0000000321 00000 n 
+0000000515 00000 n 
+0000000709 00000 n 
+0000000903 00000 n 
+0000000971 00000 n 
+0000001251 00000 n 
+0000001322 00000 n 
+0000001904 00000 n 
+0000003077 00000 n 
+trailer
+<<
+/ID 
+[<a7ddb7f6a35ec00e5bf9fed0eaa8c41c><a7ddb7f6a35ec00e5bf9fed0eaa8c41c>]
+% ReportLab generated PDF document -- digest (opensource)
+
+/Info 8 0 R
+/Root 7 0 R
+/Size 13
+>>
+startxref
+4100
+%%EOF

pdf_table_pull_lite-1.1.0/tests/fixtures/scanned.pdf

@@ -0,0 +1,79 @@
+%PDF-1.3
+%���� ReportLab Generated PDF document (opensource)
+1 0 obj
+<<
+/F1 2 0 R
+>>
+endobj
+2 0 obj
+<<
+/BaseFont /Helvetica /Encoding /WinAnsiEncoding /Name /F1 /Subtype /Type1 /Type /Font
+>>
+endobj
+3 0 obj
+<<
+/BitsPerComponent 8 /ColorSpace /DeviceRGB /Filter [ /ASCII85Decode /FlateDecode ] /Height 1100 /Length 16799 /Subtype /Image 
+  /Type /XObject /Width 850
+>>
+stream
+Gb"-VH*iUnr]fls.5DDO18`G=S^8g]UQH7J-qQiWV##6a'q4>7e`gaH+[T$?d99!$$nW+/U7rm[2ma!/ZZOdtIFc,^k?6-aS'C2bqY#76Z.FHmq:YN(F1ndS^7Ro--3+#Gzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz!5Pf.c[Yq8p$1'LIItpbT74@ce#"@<2urG0o#i]$Zu=hcJ,]AdB.6re@/9BT0=g/,iPUF`A7Q\T7\Vh+UrUUW!H[DIC"TXuhRn.6OJP3n!Za=D=LW0RqD*Eklo@t!a7%CQq=CB@*PXO1'g$X-]\b:\8G2lZqsXR[a$8F(gMO4b5rrt8?R)SteQ6*`\od0A*dF?-s8Munp$:,,dn(pX'OC;ff3GY-oB+h0O)d0g`/&34]lq.`niO<qeZ*B7W`?#rjf,/LRPA<l7j!0/]mKLjKaV72fk"BsDN+gKrqY_`]m:,)0YdsG_e-f@%F733>lQA&r/)7Ag8IlD-_>Q,RI^@Sn)q"EJQKTWI<Fi)l-cDde-?sr8a8D9o[DEp?(79"26(62QML+5HfU;ZEpnUDe^$>G4K4Cs%37L:=Ku>q`f3VcM8jVoIJ:_!O$EV'?ERVJ3\o[3%mMW<EC;?\Z6TXe4aZlb^_V1QePGH4NY6\5IGq61^,kr:;pRMjX;D>cbB<.%h0.Oi<NB/[Z#HO3Pra'Q'RfWAh7H1;*#[;njHEC-f<8P_&kb[pI-@8Z@Y<'t2eg]Nd\KGA*F[$N+jhR@>s)9(I/.j04VqraRT\O^I7oeQdB$KWQX4teAn>L`]^WcGp$1'X=gDT*g9h.o)"@.C5::6[rV#!J>.&\G_$A/M07W#A2t#J31YULFdA,#NpY=cIMauP!lX0[\=r7ar^/Ej4hgK^h*Hso:.p]aYKbn(j_$;&MHeW?/RlY]Wjd0=\[VXW1l8<A.m,,SXp?gTjV*<h9Z*2XndlVdAI,]u1IeMH-d=F2ohRm+(EmaJRS2g&,B:at<GNkrj7ShjPGOOa@UL.odU;b?E1Uh5Ueui;jW-H"%onm'>[N*!ML_1kle@Xc)Z#o6Y;bf?LEokGlqXg1;GODV>]@,lF>.#i3>eu!qcT'sDkT'?`fb7)\E8eYmhgG#Wn%9=76rUbhat'r;fb+ZWs8;I9S8^eOg),7Npu74Z7n6C7Kqf6oo#`KWh035uc8XR$;l%5SoNq>fg2#u3.bD1MisHMN7.hKTAk^^`BPE)'%>m-3XnjVOYTO^>TuZZ]_^A-F&4*QlOWjl4%a:16J>:,u;I=\.M'n/r/c+F]+%D=HZt\"QC/_HJ$sTuG`T>:0B=l4ShL"][bSrZhE3`S87RkFY*MX&Bpu@Ejm2^2rRQc.GUIL6'4ndMjH1F@1))6'XUe.PHDku:L?+bE%FJ.I$F\@G+jAOs+Dn_X-Q26M*rqMgeIQ<hhh07c,W`:RXQ<[Jt_hJVYmX+6KWJCF(nf,$he/+)rC>.=FA!<66P\Ag[m:iJh(<n^@oh2GG03l"@=Y^3);)Vp_&oCb9"-+#k8psS2*e4+X8:?[L:JROrA]pAVeR!b(oju:1M6UoDh2>;$]'HL9r(`bAEQPsXq;db$b*DkjZd0-MgYsUrNF^G2758Pepu74,Flgl;[p?Uh8Z.uJqBE@:/=\J,8nu=fldf(:<pc2WDM]\V*HA&VV.Rr=Oj&\0)M.C_`u[m]\sQBp;Pi$hOa3/g2W[V\0dMU7GP9^&DqqiW[9+]i&g^/_qsM)nHEJo%O-.MnET:fbXSM$Ds*<A-D,MHpPfP0DFmn0MU>>bCoAaE#[LO<=,bOI`cFS&Xb,%atN`kUh07Js'4AVP2J,"5A$5NEqK+F!hqr?o5dn`3aF3cuQ$Pn-X\`YaJ4nT7&1aq)7]Qs*H$ig"^Nt!!aB[s+_FMs=<b*FUuK8sKnL_1m&jia!meO;QY#4-sc*4VN<Du97q(L;.1Kbn&4aXspQ-_Gc!WUGg*hT2#ZG4"!@I/*1=0"@QFOWmXMA,I:(-0n0O^enQJW@p<cL(,)nF3Ze82Jr<go?SpeW`6`^rMXGRQS770HhW_"4rN0ocICIF;l:NkC91[NEneA#-#C>;aTJ1"2dmBWIJ7$lRPh:`rqaig:OPTb%#H#]q.staam.EWGO3s:\^jib>e*i!Hi*3>M\e$1B$?WYXmX!>E"Xr#QMR?,7AET"?turjhgMt9Z!sIuO3X'fSpg=c_M&@f%K,_5%jU68r:A5NAs)q?UW6\>?\Gce!sU"]/]Z&$?@"%iY7i6V8Bnj0Q<QtH.Q(kaS65`YCT^5l\bFTP:ISoeIT-%7f3c;\]'IqK=`\iF(.X^)-b%rYC\3nPa,a#g^3K2-MA7Xbn(p3VO%'b]YBTHldHAfD2(KTt^%^BQl6W3&7C5]+A@Bu(^Vn=qq4I[Y/c<R(?aOOr-IlSa5ibUenOpWGrM$0c*]Ag4G0`>G`*s)j/kiHt]6<SQief7$eZ-$+rVC'HW2MFG\_r&I[VjnNo`7R>,k[EO;l3?_5#5g:nA,>K1h_t8J,YpT\gb1n0m;p9U.'u0LDr*)#-dVjFPi2e@n7Q>hgG%?jia!?o?B5'Ct\C0='nCH4T"KOkKTJ4_rOoSB#S.<rVd);:U=][NbfIOkhUI?4`/o[BH]%Dr,V2Y[A3+`T0>F0?Vg)qGi=?4B?o;HI=(McnA##%n*tbc\oZ.WkDV4B=Ku=(ArbBMF6Cgc=[Yl0p!m,"1,>.K?f&/(dg2$I.?EO/k-ns+TB>q>b:fnsXW!TZRcTPM4(9%Rie]&>E5N.,pB@CnhuE]Es2OLB.TAq[98F`bLCP>b4ndNSp$2MF_[fWeE0Nmk?g;@(-rX!l0#%,?;56UW5!Bk?DRXN>+pUTiMk?+<UO.tm(h;E)IJim1d?R3-S2=qXB%AnAeZ;n0:S'\RWT0"#0#kag597u:;m_Wi[;$_'bV?(emqU@NOa?.GL'c8dk0'*pDV_ml=Wt"DRl9dl(5l1ee,8je:Qm*2V7s0_rVC2AZ#JR4H"5gcb=<Y&FP,Im^MWlE^:oXmr\tg2[&[kt;:D\]p=`&WafP."[qsa'qXs.rJ,Jh6-OUY=e@BY`7j!/dFQq5;-Vp')JDZk+)`Gp24#d#qUa0P_A>HpM/Ri.1];^S[ig6K7Qd20c48FVD8;U[AX097+3-!r2#ns%sc2#0<VG3Q.^ZQuueXSX@<ifDC+Pl.'hgIG)k00)@=*fmkK.C-To&\&i.H[^Gl.%G97U)EJIt$t-P#3]=D@C0<^_$YKzzzzzzz^i.\r7>pTJqsLMf\uK8\Nh]('b/hXpc!%l2IG_be>IA3gdnMbqT0E4Gs86MEetu02"qDX*[VadMX]tPghuETB^JEbLB:jV'[;4AEY.12rNuiB*c:qY#S(QQ$@Dgg.Dr7EdRit.Rm+f--gB,!1mHW/ucQeK5ZJRSrHhQfDh7C"1fs5hjhJ:`G9hjNYHhZ0tA)f/ZX>le2[f(*$6Y+W@8L7#qCt>]@eu`-eap*@kjd2cc9q42JI/3=H-;6BIS)sBV?ZkM_Q,d]]=`V7\&E"[g3`E+^Z24.Zhf6[WQX5"&#q)7i\T?s$gUG5)'e<>7WDf^<mFld@H06K?>S;t.r:8%Q>$laSAbG?=.o`Hdo?B3a0e?8#9[0"`QS2^;bT/lpGV%XS@Df%0N>aZ(;P_Cl_1VJc&pN]p&R-Yf[VYlSn\uM@Pq!O!Dqh^;W<Hb535S"9$=._t27^Ja&L0c`lHPeWn(G$[*8B_UT>"jXP*d:pM^+0jLq8#hD7@L=M.7JcFQgR9j7/auKGJ'KW`?)h4T"3':]AP>7GEW^IdOPpDVi#d?7oKm0UgA$H1MNcOn/CTY<):@b%!??/@jX>#]L?'o?B5EW`=sDp[c]qB4dFC!<WN?04'%c`//j7m"l4`?G1[,7C*hm7U:m"[;+6lQk^js;boQ!n`%O0G37#:'E*1MmbM@\?[r#sKWZ(B=sOo*ioam>>o$Ae0>=9A*BWtjP>6Gm^%[\V0pI^0pZ@r`-8/l,!s];e>-qt74F(sCIBi;jWc8$NGOLQYX&lL-;56U6]bnMF))J-XM%VJUpgOjt,9^<>n\o.3F]Pr/P8jeS2JUdBp%6^/a+tkLd0P='c.6@sX0986]^sEA;BPftql2HeXup=NGj\5e?2t!9Y)$p,cea0OrR@'7ZY#bQiebuEpY;2pCSEZYmFsRBNGcn'.8c4^%NQHF0Y6WJm^k.&CN%n3V/bP80c&WVF6Bc9_]iY\@8(MoL:U<]:<beLNI#jJp")-;e0rj_Dqqkmm'G#UgU:sP\(Gql/mVpkI/j0>a2Hb3hoLp)UNk(S3AYpua*c2rpJe7uorG:1HL8,l>.&+1b+(3PO[gILKi=,!_^=),1ob-1QM'\%48RFi<E$4>/6L8r9q*"VAgB$ndhFN#1[>8D9/!tXVbWdr%$`>E^?(YbibZWQ39`0fjHEC-f<8POkW#8!HhL^OBMkl.o#i[Np=Nm8`f5nB'92_Vb:eqD+.;lId4L\GoBXP/[dE7`AF9Wjd(O"@P\6V:2As_6Sif#*e!Mk";R7R[<N9"IdT"M=[@_sJH%cS?HM$D5FmRC8mu/a*FQq3c,=Y)'HfprDP*'c[5Q("<PdhZM\07+YUXhk#Mht=0@P9IuN@k2&dnbJie":YPIBhj=GMNuqrV"ua;\LDB`>eX:UQ;]4*8r\'\SHVK?UdLSldi2ur5LD=Vb\sM[VVJ/eD06p9:%7oG3mlBAMf8-qp%A!h"X%$6$18?4*M`1If,C*5$>$n\LXkrEmpm[&KT&]042"qaZ,t+pcnQ3KBSHTHgeXMpZ@r`/lSW@"NuU]YJ0kuH0:`"X&c@+^]+)N]Q9*B#@;$uIoUd<A]Z>d\J2i3qCVgB@d`$^<N9#4e##i-UZC+@8WtKtk!l4/mBk>'G3kU```uKdQ7Z<AX]kCU0><6FmFns0j,>d'!*c0cM<#hkHM-PT>duKNmT-Mt*P:XjZY-:5`p0i-fLpO>gM\rEme#&3>Bdu>%B52Z7uo9-\8_Ai[cGn5595OO_@&$pB7Cj!CXG9(FMIIr9:8rWfqX6XDr*U%6n398W@uuF*I$9N%4Om5nT*^3OUjYN?u."@*StV3Y<[X?WB4[tB$H2E)"I&#]PW*Bp020Y?$XY*ad"AX*_6;73#&*"N+u;+If8>jaD7um(<%d%>1qW;CEq-4Dr3alXMpP`#,fl'-N3V/4aX?ClDcslnhb.5m(;`#M^_NV,*/:ErqSeXpu=<jHo#m.d+>SAG^7Do,o%q29HP:qQ90<YI**n9UgKG0=cKb)B?-#QZ.!LkjEKI+VaupoOUtDh@N6DNY@u=X;.e9ALr\I/E8\OSPq.@[jd2B<iN6J%IZ_S2baF[=c/i,9UpUKt0sVWSCfo,a;Tn-aqP"_,=gKGX1)eB9Vk/57rV%DZpWIJJm]#[,Q,h[H>deJ%JJ:A"K.FK];RUDp-;Br)3m+qgH7OkQqD0^#Le^/0j=X`LUrqE>V+[Rj:4F0Bl%iigO$EV'[Gp*Cp:$jc72!e95Q;h26ta2HdA#i_Dnl8EoV0s??+V"4C8?(A3NI/.m6mQ4d%NhlE_p5`L_16g;?upecTcXBhEA`n^%^CL]+GOf/R,YgE5;_Ro&VB.]Qf8aEA$hEc.fjSn)WiuP59@hg,3E^O;L(?bL3n?j2mK4a'-$O^@[oihgYIil)1-ph<PZ!p\Y6^/+rl+o!mR'=0GqY+9`>7+POMUpZ%<!?DL,S)fE//i_sUg-_GcaW`5tL?[k3rA\is%;>LJ*(m)D#3d#2&.TI[.c(@3b[7lgg#Zp-1Q&A%sRQ+p'R+VE%dA+0fgU:S1MQa>(J9U%nN=?d_B\g7BlI"_\gUB+WeuWT7VWj-Ck,cSU[^0VmT)7uk^N/GY.p&lI/b@Ij:R*A--M_Djkl$u0Dk;?>2\nSZIl\(po]agLe4&+/q(KX?%BEjNIUou7RjF-*Vs_VrdVIOAA]'63dl4RMh7CFT#?Rs,/^iJ!R5=T,qGMP8V*=TgYGptlQcB221,aA""VBU"M>f5E>1''5qt0+drG;>3(i7d'<NC3(cDc[XAB9kr<Ki>c1M4k0=L$jS(L7^qEOK_un\uMF,ZBb'71S\=[r9BO-)9AD\bB<H^%KT*cV30?=MB0.N.V*#QbT=\&Ubct=cXXh5'E(2$lKp#D;)@R:pDImV3bL,7lu"ejT"G)_ncsuoA4i_NEEo2fs:87cThGW1mR,fpGm<W6nB<;E*tFORRAuk8:Al:++<e/?G1[PfWcZh'eQY7n_u47n3XLSBGK^%HZ%7P97?^sEn6Xc>ojMqaTF)E!!c/M\#;O=5<e\_1OoFk_1MtG>duIZCLtYQH2I!<j7hc@ldi2mU.'uWgU:tU]X^Ut-7)9KGj(iVrgkGtgo,%)f<1)Beu2GUrquSs@MH<8mUO5LNaC`Gb*=L/UIU@^jMh7,\o`d_Z[MK+5Z1r?F28"r7*JlGEb3bKaIMe7i_:`fiqVPpDr89WqOtXsg1rqNnA@QkCU+K.J,apRM6!ho3nr\ZdAZ&H[r'kXVFPZ"5%tsK5Pi*]=Kl2LjO5*rk8#RK/0a*!SNC>dkU7\K9pma_8U':>Q]KV/A]kjgO8h\p3qM[[@qO;A'(k?Os#F8!k.cYaqc`PU$6[,M&LHYW#9Q(jQt--bmpaFLO$+8"[r1$U[;4@q"U^JD<@WGUDVTKUMA<DUZt\!`(]Hc3&I"A:AtD27HM-RRO$33hkah>,bLa+8e+A!jfoG$N2r5re[VQANM\ITQ4a?g)*$c">ojq``6o7eK*,!BV,pP)h=,18@8.IWj23s*AHnbPWm,,<mj7`?LU$UL@le6-XX\b]Xo%f/M"%G2lP:'/_]Qou*D]["Pc?i6GUt;SEVrNe!/m;5uMKAOA;bn]@:<&>:T0+;WC!n9+O8h\X3qKDq@qO;A%N*_Ie_Y=Lnj!O)^gRmF.10&HX`N[Nc$#-mGHpYMH1N@VQk1;7IfBe19q+%8Sih;@PUTISs-oe.aN0`)B20H6^-X/:\N#3L(+b8"POn=fLi<q6UE*sYBY8cjp@[b'HX)cPn%d+VmRnV&!XJc+zzzzzzz^NGGM/\:iI]<Ag2WfA)]NZ'c),9nEjC]4,&IJ`_TeZ..o2r-<=qXf::o:N!bRl5+%3,i9[[V]6XDV`'Ws6m`&5(/omIJ-gqY$J6@ieed?HH_pA2grY]@X#_E%;Xk8I@E-3L("s*F6Chb$gcTTX&#?OO$*!SIc&65fsA;5ldi@QbMKpZY?eau21>Aqm6cX=NEoe1br(PHrqO'Ilh7q>%^$Bg7MICUnG"O<$5=mkZK*,-pr>@rdD-6&ona>0Q7lTo_Lr6COF$We=g)/Fn%S\e:`hL.'>&JWIeVZBUL$B_Hg%/YeuW#<5'PkrH$A%`7uJ-7DSGr/ace+Po#i\sChd=r.[NI/E&^;5.TWWQrVH2^Z<_e@V'hdP6KcXb,::C2'e<>7WDf^<mFld@H06K?>S;t.r:8%Q>$laSAbG?=.o`GYPq]LQR5/>geuZE0q=8(anDV8piJ%d\7[S#AW)8[$70X8D][oKZA&jV^6J8YUj6)C\hH6dWWQ8]+[V\Yh&8T5s(+N93!aG)2,]Z>ThRn.*8-O!LVk8G7OsNb1qW?%l,FTB">4gNaYJ'o.gT+;Q'!ieG*gFD\#:P1PrquSeWPec[Jp4T#\om?$c'pZ;`VfQ'pUg8,++/%eM'p%3hL!WkU7p/:;+'X/TSQHXN(d_P'%JYhB@NYS:,unE_M-+eBQ>CkP_`1+Y?nmp]^Wcd6r,r8`4gi\^5>oCKbWHaqW=8cW`?)82<s@sJ'Q9<=p"s;-sjR]`/,0%m+E&)h^E]j7C/qP&^6u1?F=NJb_d"uWMc?!Gk'd(4*0Q..:9f:p[6jI$X\D^FbHn7)K4HC2.jhRb:^CXVPIRH9,Tg7lq&Rq'W![43GR'B88`@[q.B1QFujU#^^+b%o&RnUou4qGW5\kVYWZ<q?G/CO6.GB@HGB$n.hOjAAN&h!&ls&qZZC(SlAL\qSTip^9o1Y@N@dU;9q+&#A=i\`;BL\8KF&;9L->+K,>m%OjamXD@URi^6\d'11&-_VGh-#1-&ld)cCIgd+qmETbdj9m'rtRHl/?rH!<6EE[YSSb^]/atb,eaPU3h=V\`_PF=VIX2Dr8:.CNNLP9Sl:uFQh)-LsbhNGMdhJi3[(=QH"n'7tlo=)#\f?N[$K[[M_HgWDf^l?boB'4hUdjlPh]rPj7m"DVTL1fk7d2gGU6ACqNE-/l@`!^]4;#R!\*;3HAZ+a-e_r-84DHn%%VQW`?+&>e+n/ZtYRcrqI!gr/eXNb:gUmrC/]#>0bWaRFMDJDr89JiPLd*9hJ_Us8;J.G%05=NA/dgKY=2Wm47k3>IX%/gMVo_*PH<d8sMG[l4ee6g%^:2)e'Km';GlQLu=\6SoMcl-CRl:?G.giXrK;K2'kl8Br@L;"n9PSIr*K)6)mu(XG7mH\4i;G3Nl3@;QOmfUkdtN!p&Y"B:ho$ohD#?3D.bf0k4+fU+f%Y)C7-$^"Udss1C[P$9`_-L`gp)STdPg>Inq@L_1l+ejW*$>Y*'fP#t]))fE//Po_1:[;+5A-:TnO]tF1'S$D(IIer;)J,3(rW@sRMl`Rj/Vr)t#D//'GYIt]ijJ.c"b-kgpgR/!$P>?M3k&B$il3f?.KG.,EhnO[=5%r<98:WYTFAo!I\))E];IB7jjc^a/YGj2/<V,@tbEXS:@).;+Z'"@HIE'VYVFNS9goZK;j2[4qq<"/Vb7Q1a^cU^-(7"(&)fL[$62L2L!s]<pdMG9lUma_&8k(+(!:/4<gS@E)W(>UMP=r1WNs[>OiA>cHgaZ=(Zt\!$gU:t@Z[#-Q915Y7+`MP#g$XhhTki>eA_!e\^\HfHR+8?F]/._lUJ[dU,taDXEZ,mr=0Grq3LOui=0IY?p$C?f<W-D"fRE054hpACou1*9?[[&2j:*%dUaD*C4ZYY#pmJdR>s/,]j41/0bF_$!D7A>e4LiI@Z`hjp9ePF1Y?nmdVa[,fp=a:\@t`/lq^ae`='g.uF!%B"eZ)W3%+DOfUrS?-.kY2Cr;5rtY>_>oSEQb_`ASfepu+1:r@I2rZ4K`G`f(eWPasm[[Vad-8<G)fl&4O7_iP_&S:'nL0epgSlDicKjXaF;?:Y),Vcg?<74kNH^])hlJ"<!MXXo?0qR*W:CY)%GCYujSHEl26EFZ%MXeKWl*BQ/8l_a&L='_(=:S)1CRO.^d<j#[FH0kJ#ci<OU+2@X6S"W3)^6W?i`/.-Q/;ZPOIEWX7mbGAE.Ar]!*((o_+jIHd*BSG_:6R2]WF[8]7$SfREb+`]p[6kBn:M68(n6C>(rLVFpMehflC0SB:,JQSm^aosiiZS!Q7cJ>f>%;B>%!l"bQE*mDp+Fu1M>"Grqbs(PUWqH_1L_[XAYlep#se5o&RnZ55PsAm`[K5N9W-Hr:h+B7"^>.'+HU`>a1%1K*E&B=u8^V0qda[WDead]6<SQYF?>rHRi,%=@`H&S*3Eg*L/3\[RlZ0WGaA5l`IXj=J\7Q(J-.^]Nr*j8JS;iRa4.kh"6Y;,a[=r`K*7_347c'rq<gsmR=%G5C`YsA3'+ko+io]FR#j'hL>@hl)Cjl]6;m(oULUlgNFc:G3I;HG4"#Jh]=R_`/#hnW<5eML_pN?0h4.sn,IJLTREZNkZ]:%8&3<l]&Om;]R'89a(;5R[;(6o"d3$c8q#)]-YG5RhRmj>.p&lY\od2tVJnAH]C3JGI/3=f#)VIYUPWgEa8q,N^A-]nb:^u)0AeM)*7LQHK?#$[XB`;Ve##itl-Z4L#:?&T$a7569!kGUqX\%T/63!R8-'3i-tnZ+rcn(J2r&@>`2D,W:HnS"`k8Lird&<O`2)H[WT<0jWDfisW=i>arRp>sPe2/.*]5bPX&6lq)&(aH4aQb/D;/N1rDf.#n8@$k0ktOjeu_k>Fo1n"c'gM*g9bSjBKBcLc*j0+HhQ6Es8>fe=0>er>3;f^Sgm[99Vd924*Ri7^HT?H7NdZ:R/YYreFo@-U:!,Q^]V[q+jOfFQ7\k_7;JDLB21:6H1U1;)9aE"IK''*Pq#e6?bcU0iPV!8?QQ?5aX<^>/Yk>H3:G74,Q>Yg0"BhkrI.k9hlX"b[jHggKbn'OaqhB2$:T!_0==IV[MYs_#\%.l/SD?BD]o9DCnc)A-X3rkd6<!Tm+SitI9-!?[;+6LqNd"L6\Z#+Si=392e0jdMA7XJFO<N;^[_)Y2r9@LQhpE@kbe7Nm9?l,am%9PkM5s:\))GnR:iDbk=4>LUYrobjY,lE?g[Tu^+b$6pt/%#.DCb"N&GP,6lu-D5Q280S9EV#WUEDLT&_>"W)0R9^jda2>C0(m1M>"7:n^5Ik05.2/R,ZbI/3?4-h/suY3<l>MN\cRp6thaHKh8>BbMge8=#>+J'0FQ6k:61n_7usmF\[Ro\.Eg==Y:iCk(]tV2+YJF7fA%dn`&Jl4b*hGo:2_g2!n*pu73A(c$!a\4D$Fcr(6)?G&7$;(`s/aB>Tb0;JT';hVWY5A-O0ehbW7oCm_P4F$A4KaX*O<i/QY\)2WrIZI77DV_lU>-s)p\9[].-;Bqn-rJutLE?;=Rcp!2ZE7EE2DuENo)/=TA&Tj6Pq,pr(&.PY9H6=.EohW>kcoR5p/mt\QS,O^[Z>kW8Eu+E0"BhkLj:LPY)G%PR5<]$]W',V.ub-V]6<Q[A,2Mbm^qpWh/;#XnYT$M%dkF&>U84T*.?_n=e9ltM,p`^7:rT!VbWDnU=iMH<\4qlS>OAGd;5P\4Q`#@HM-R2YLGLSm+=;(K*DdBM:$S-Qft1P(G>XELEAu!NDggTc'gL?:3q$5B4jecne4?SW@/&>gUE4^&e4<[='mO`[VR^*=MK<5[$g40X+0irB:2oumAd4i.UBW-nt0\cHY_36ft;V%d^POK=QNFO&+_jdIcmF.4o!rfdBW]0ho1i#*d]I#7fOE9WBc"9V+R#1T@&WFT3n'GD-D%bJ&ie-@.MRd;&,j**dd7Y`JYPYQS2^pVG5P.]k0?\(eR;dNUAk`5a:0uY;f&m8F"CBH8m9M?h1s['eEH@d__"b4#4-[@6aMAF7C"i&&duXGMI2?>IX;@WJ=WP^1\$K43<[6:Rm>hp$259f.T:A795I%[8<0&#m1DB]^j2qcT10Y%3%4l2>+pZTj,!D&LO81`m4hu)FTot[;+6=m^X^.?M)aPMeL>Vo3.8"5Q16',PMH$-l"=Ke>Q4lSTIc:aV`,h+aWe.7oM?1(+`Q!<@+)1'Ta^`0d[o(HLsG.`)nnOa##RG;69Ik8-+`/;;@m2N?:B6mbGA+D?8oQ(628]KF&;-04)<KLR'l9g,+*Gj=+$)jSU4's%#UK`gf&1ib3CfbUm&-I.GB,;52[F?0gf<N`m%U(+rk%FQq5LZ0g5N-`5t]INfFpj'9'2?#j,D%A`tRk?.8TYJ,l[8o-3WT,1n=U7"`H3q`IGhp;,>9!k*WM\a9fA*2"'a[#Dupsk'4zzzzzzz!!%IL(+i^3pYC%#Fk'J44*DT$C/dYl7V^NNp$;MKmG"X@X.+"<C&7Jdaa\9%fX,AA^juUB[Vad1FmK&*qg\YEpZ^H1m'G$LG3sUp'j(%kd:;V"\`kVV'>#2ACfpT.mG#!^1M"RgSND%->@?'c?%)SAH/t)XpA7<kc^rsp>ITnM>P@L@HgTOpHehc9SU9h=2s4L,UgAH=`R:SU:S'[/I:BR@If9+CXB9KAr:u+]^Ti](Iei(:D4LW!p021dJ,fIjmFn[EYp+H29l%b-),Y@IEo)*)S[2qQg?VoM(+b82hC")@,=h1D11J1Xo]`XMil-fdaZBXHb`&2pXB;b(p$:LT`uULQHKb_`SC"ffkO%$5a/l4XFm@JU?+Y:0ZqE$^,Skb&jd21=^ARQFlIDq:W)2I8Rr==g=Jec[RPek5?bSXj4+-hm=L_eWL(PaO2/6*d8<L`@#OaPFmF\Zg[SXDKg%]jQNAm.Q<@["c[Vabs$cH4gKS9aJ]&oGsEdo<24Ce'Hc("Iofe\ZUo[?TCZ=HP8A_q[lNIh<A??kF_l+>k-GlIO%M\[m*C!rgLM/dPXIe_l84S.^[3tsb^\o_Wp=D#jKGpkSrs!@QE1JurQ39,r7\KiDa(eQ&J3pPI*GMRE@q<(85Hf9YpE`T^fdRKcYT;G#5'#tu59GG(o]6A2GYi2[icH;bDJ$FK[\So1SeffO\_!f]1hgAA)>.#c#5QCY&Ale.80OR09il*9?S/?n*Y$JZjX&gsbBU4dC&]$#me>cNL@b+74h/qq]Vl%J0F6:]'C=^(FpXm@7?[j)iX$qp^>ISL0=n2*hb:\]%HbK24G3mkF:-.XQpYJPLNZ'c)P1uIE^6uWmSIQW!0M6,P4F(sCIBi<5Vf;]68E8gseZ3K8W"F#+`p]m<ib6Y#o?>jUe9C[.Ojk0fmYN<_3.:Yq6a:Y1ce*0,M/q"@9q,fhWRSCnfs3)-Wl2b/DN0u[5@_>D<GpX(Fh-sVe##hU?[Qi=]r4fuEoFlh;#oW`a!#\3!80dPMKnjoqsUG[baLC>MVE3u4h#%JR`K`%qt<aD?DgJb&s7"od%O=rm^V;;qsM'LFmj.iQ7\(82nnq496]gZ)-q(f=!KZ$[^fkdMQ'OPjpYFU*j<?XqA:c%hRkk:Z>npEZd1hID;2K2hI>MN\o`eFn,<"<*tr]kI3V7BN8g8:WY<F#cTV#%X]fl/b*C)5fL:J>#d.$FV7sk9!)m5M[*(,;4SmXm;K'5<`mj6WY[Bb=]/+'C8?M[.,EG76:L;><4ZhA(1sTdGc'gNepu@DkZ^rmbhnFN46A<!nlh8."+(Su+mOu6B64@BsQ7Nn`KG`XC#!4KQr:4?Y4Z735\7fT]SN)AVj'hc$E.mVfc6QoYR+8fmn$6NJ-E6IgDng_[M'o%VV3\[hm^luiqrO/Ok-D0=CNnW\B6niNqsCk&S"G*!aWYnA!;lNWZd0uMqBf"0*8TM>0k4+f,.ppU<%66pcFXpcJ,X_I$5ImZkRg*5b$)a&V+[/8(bm4$j#c_BPOtt;rU@]8;+ki;iJ(<XN2kO1<NDK(nLsV,k*p;/Q@JbqVa(.E:V!:8=RrW,mY[Q).8ETbmBtR*NI:BG8"!NM1=AUD"q=V8=qsDFWUK=L;W%9>j"o-%(u^#N*_<-A3*GcuMumI!n]EAce?)q6Z<u&%`eYk6i7P;EPH@9+M4qbrkFNX-a#T,X+"-g-/28k_BaeW`iFkd&++O4akFXtPd%G%Sd;'ON>W@mr[)kL[qF!`ao#`Im^_0k0M6ZA]eZ)'*I.c#kcTCU.>.&*nrUnc4n1Vf?Z/7SUOmKI(:4)HCpL;4*8q#A\\Z8l>js?@!P2,!JQPfB`jQkL*B6$WhFrftSEokIBqXob%(t5BWS!M5Bci==&gpqL$T5E&S9n]8%ft?s81QeZ2qK;/W1#(aY;(Z"\Cg"/\eM"Z3,%e5ufC"SRcA91rl<EZ0[::HSlIB+KE+.&JX^8\%Fr.P=<TmobfRE054hpACou1*9?[['MKFJkYUaD*C4Z^1Lp_ghrnA,@-S8uc-++>%u6g?`W,5('BS+dG?'Bq#Fhi8^=>-qs?VbWe$^A0*YA_FA#ERq;)m^hU+F%Le"E[.fbY#d8ZZD;jAma6l(rqd_maFQ'"l.2t@lDB(%6u2Gi8a7g&TKqeiJm#El`ASt?*QGi*8(_/6b]hrmf<A]$HgeX,)W>"$hpeW(@ouU,:n_sU3@FK-o3Oj.]_LKEX&lKr)GS..>ISL@-Vp<>Idp'i2$t*M(elaZkQl-(6r+V=X+($D8Psg$[F_-K9WjX,eui9tENX`ep?APpj2R(\>.'7%2<g7tm+HI1<EDU0CY#Ri'[2rMEokHb@o4(mr:.hkd"`)OqHB2A'.66S?*RkXoS\L0jj'DY7phm;R3cK'cF08L%'e:B$=R\;GM[VC8<G)@7uiOE)$eO#GOILLWUEB7e9Ab9-U]$QqnJ&b(ND'*?_!u13+Qe5d/uik)%A$]e/.'t<if<%errln\S0BboZ!;jqVq74,tl7$;KLHl>?@q1[9IOZ,s'r9;7*nlCXsc;U4Jk]lDaO8J,eL'j)?ir=_FI6FQq5MET>&'r:!(dghZ2E3HJf)5(0._7^\(&Zt\!dI;OJNKF(LQT6T@P<"%TtIJ^nm5MW#(O(>A[rlQap'J3LqJ,fMNp]6.F<1$3H;:(agBp.cVhKnLiW);f?r'_5[q;da7%&Q):V+[.Ec^XoRN9%9KXI)+SQ2'+krg!].87MDdk>/8M`5";-kic5h^]!luNkSf2Fu=$a8.oDVB/RuCS]I!83cq<+]6=%[a,Cii,3n?i;5:;/rq^s6lHPn`ql?p7MO`A,C"(_gkF[7<A]pBlr-)]/pA8*%S<7jRb!g>DT,Y>A;Pd!d<isLjojq_uf@R<>b<gmi`Gd'=gVA@BP@s72;r=66K<foA,aqG["q>%cZ>8'p8:>-AF]PO@I/E<?Vk8Ea7*n<s-fO9PZd-5R`Z*;P?[r%IYHmffgUC.];WGcV9<V)>0`m?mpXE,;m<.GDZFK>6lRGu=*&TO=A(\G>Et"-TB<"_B+VNo#n])`4H?I+72afltRf:DVT1"NoMO0d;@)7FcL'%Vg87#&^3jhSF.Zq-e#J\YbGOOBO8nB9ZeAO(O`l.R@NukuOgU7hI<[8u5mFnre2f=Rl[rq#fa\sg[=bGUDc'h'T97IX0B/RuCSNV>n6p-9_RX4`n7/;gK5/u4)X&mVm=#"XM]KS8u3:/Zt`MnfHpQrYm-%-%16]C)Gc/b'Vao2([XKPZ_R+JQN_t:'"X,'&6M\[mR7$)d'.6[>5))A!UM%]'*Om'EPD05-s0tGjOWA!ptJCO<TPc`4VbqI'K3NL23QX(Ad/]J(\?i(`$jVS]P+<ns*cAflc-,q/a<([*jJ]t3b,O&-9K9onjBoGrRX,?r_=&cVaI*iu/l-h"a3Jp#*<ioPr4*OFgZKO6B%o)oF))Gl(XB>"eM_<0J9a*7:Nf9<N6n17Tk^lHn/=H);1h,p_hKnLUB:bO&T=]LHnu/)C?%:u;4P(ChNuoIU8>UosN/TZ3g=3<5ru_:9a-U.^3WEGuc^l0;ka3F3T?Fnb%_DV9^>re^EGADM?2X965Q:299.l(+^mmE+8s_.nE8\NB<]$1W2G*32nA>d3D(Jo"/=@6Nopg*KC#6LJ["72RLIYC82JlW0ecQ?Nqn;a\S%;K\H1L%U`JY6F1aoo"/G8tejfC8c%-FI5i.2--a@'Tt#T(IaE8e[]^etP1'92Q""`:I#$NBJ*n&FpJs81^&-Uqf"?bUop]i7g%5+6TTPuji4<Be47'&b2f<V1L]4S)83;6.Nc#;9>l0M-'-,JX3Ng[TX_WcdXYD:,(H%!":FhV=nuP7f064S6ufl-F%VpYL6/&_F0k)]PkUUINbT=><]_jiWiu`P$_`2\)Kc\T0iZoSB)%(5TeL/=c+&*s[u*@[RTV\0*QEb*=KPl:,?FpD%TrgL?c?4&QRLRp03Fl(rt8$[1gb.>AHIq7:_7(;ko9_hM`7h2P,m<uZ!$L'DML[Z)Ss41rso+qS()h0&&8YE.f^b=B.UqsM(+j5\.=M_2ji@eK$:dD2p!qW`l^d88lt"Uc)63'?5[1KLSZ+n^tC!-KXe\\8o)7#@sGY?pd"/Q$\[2Jj?iBBn!GR9>Rjr;"Zg:>JX#VYKAMO%.ksk,.LtYp4E.8TNn5OjOH0c$/Bu+o[nBo?KFQn9kAQ=&FksI.4s^N]AH@rJR'6fOKu]goXeiG4""k5_0!aNQ0;aN>j8F)pu$,du"kjQLi*d7)@s5B@EbSY?np=Z,[n)Ke?`HHr1F'n%E<pL,[t6,ALrB]',esjI?-S0o6!(IGp);^518T`u.l6,!iD=]on/_dRejWb(6A&?+Y:Qn)$?!'7$lOX*a^-Vc/qN,Mbd$Ms5NYk)GQA3A"E'B%@^W(oKcW)soOjQ?MZL,u<_nG`*m\e*js@cHNANDmX6--5MDs`5F]Zl?rHmVj[S]fSJ&O480hn^<E[Za3eLZ479TqF@;m]+j'WM^Rs;WQoK(p<3oR$8T&9BW)+Z#LO>kn[;+5AJ,]@C[>mKNpt4)S0>=sJHc,o>5npRpHK3&,T0%77[TgPXMhRoVT.j]jil-iMq]DMLg9bSbjN1>t9SpX`h]-Kd\iFio`6n!'A9.=#5e^XBJqXRKmPVAlXCtHk]>[l3]&OWFA:)`gQ<_1RIJTi#AMfS2'&OjY$iEc<O<bb*#9S>tABC"anDmYQZXuo?Lgi3SYHKhfzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz!!%QLf,bnCz#l4M3+TVQCzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz?mR_=EA@~>endstream
+endobj
+4 0 obj
+<<
+/Contents 8 0 R /MediaBox [ 0 0 612 792 ] /Parent 7 0 R /Resources <<
+/Font 1 0 R /ProcSet [ /PDF /Text /ImageB /ImageC /ImageI ] /XObject <<
+/FormXob.74e5f40f81f83975c5a0766802ede527 3 0 R
+>>
+>> /Rotate 0 /Trans <<
+
+>> 
+  /Type /Page
+>>
+endobj
+5 0 obj
+<<
+/PageMode /UseNone /Pages 7 0 R /Type /Catalog
+>>
+endobj
+6 0 obj
+<<
+/Author (anonymous) /CreationDate (D:20261104013809-06'00') /Creator (anonymous) /Keywords () /ModDate (D:20261104013809-06'00') /Producer (ReportLab PDF Library - \(opensource\)) 
+  /Subject (unspecified) /Title (untitled) /Trapped /False
+>>
+endobj
+7 0 obj
+<<
+/Count 1 /Kids [ 4 0 R ] /Type /Pages
+>>
+endobj
+8 0 obj
+<<
+/Filter [ /ASCII85Decode /FlateDecode ] /Length 130
+>>
+stream
+Gap@C3t?c`&-_qo`B/XV#V<DRX_^ZZD#HV3Lu,lXNL==7aY6\F5OTuH"k4cL2*qA6arj@p!aXkZHS9++IoAMmqcr\mM2=,.PrpGOZsn6k.WN`a*KOmAT[,G.kQ*#c)!C~>endstream
+endobj
+xref
+0 9
+0000000000 65535 f 
+0000000061 00000 n 
+0000000092 00000 n 
+0000000199 00000 n 
+0000017191 00000 n 
+0000017447 00000 n 
+0000017515 00000 n 
+0000017776 00000 n 
+0000017835 00000 n 
+trailer
+<<
+/ID 
+[<bbd50ca8212d24bc51709a8767df1587><bbd50ca8212d24bc51709a8767df1587>]
+% ReportLab generated PDF document -- digest (opensource)
+
+/Info 6 0 R
+/Root 5 0 R
+/Size 9
+>>
+startxref
+18055
+%%EOF

pdf_table_pull_lite-1.1.0/tests/test_lite.py

@@ -0,0 +1,51 @@
+import csv, io, os, sys, tempfile, unittest
+from contextlib import redirect_stdout
+HERE = os.path.dirname(__file__)
+sys.path.insert(0, os.path.join(HERE, ".."))
+import pdf_table_pull_lite as lite
+F = lambda n: os.path.join(HERE, "fixtures", n)
+
+class Lite(unittest.TestCase):
+    def test_pages(self):
+        self.assertEqual(lite.parse_pages("1,3-4", 5), [1, 3, 4])
+        self.assertEqual(lite.parse_pages(None, 2), [1, 2])
+        with self.assertRaises(ValueError): lite.parse_pages("9", 3)
+    def test_ruled_tables(self):
+        tables, scanned = lite.extract(F("ruled.pdf"))
+        self.assertEqual(scanned, [])
+        self.assertEqual(tables[0]["rows"], [["Region", "Q1", "Q2"], ["North", "1,200", "1,350"], ["South", "980", "(40)"]])
+        self.assertEqual(tables[1]["rows"][0], ["Item", "Qty"])
+    def test_no_joining_across_pages(self):
+        tables, _ = lite.extract(F("ruled.pdf"))
+        self.assertEqual(sorted({t["page"] for t in tables}), [1, 2, 3])
+    def test_text_table(self):
+        tables, _ = lite.extract(F("plain.pdf"))
+        self.assertEqual(tables[0]["rows"][0], ["Date", "Store", "Total"])
+        self.assertEqual(len(tables[0]["rows"]), 4)
+    def test_scanned_skipped(self):
+        tables, scanned = lite.extract(F("scanned.pdf"))
+        self.assertEqual((tables, scanned), ([], [1]))
+    def test_empty(self):
+        self.assertEqual(lite.extract(F("empty.pdf"), strategy="text")[0], [])
+    def test_cli_writes_csv(self):
+        with tempfile.TemporaryDirectory() as d:
+            out = io.StringIO()
+            with redirect_stdout(out): code = lite.main([F("ruled.pdf"), "-o", d, "-p", "1"])
+            self.assertEqual(code, 0)
+            self.assertEqual(sorted(os.listdir(d)), ["page1_table1.csv", "page1_table2.csv"])
+            with open(os.path.join(d, "page1_table1.csv"), encoding="utf-8-sig") as f:
+                self.assertEqual(list(csv.reader(f))[2], ["South", "980", "(40)"])
+            self.assertIn("Found 2 table(s)", out.getvalue())
+    def test_cli_scanned_message(self):
+        with tempfile.TemporaryDirectory() as d:
+            out = io.StringIO()
+            with redirect_stdout(out): lite.main([F("scanned.pdf"), "-o", d])
+            self.assertIn("text PDFs only", out.getvalue())
+    def test_help_has_paid_link(self):
+        out = io.StringIO()
+        with redirect_stdout(out), self.assertRaises(SystemExit): lite.main(["--help"])
+        self.assertIn("https://tidybyte.gumroad.com/l/hkxqxh", out.getvalue())
+    def test_missing_file(self):
+        self.assertEqual(lite.main(["/nope.pdf"]), 1)
+
+if __name__ == "__main__": unittest.main()

pdf_table_pull_lite-1.1.0/tests/test_mcp.py

@@ -0,0 +1,35 @@
+"""Starts the MCP server over stdio, lists tools, and calls extract_pdf_tables."""
+import asyncio, os, sys, unittest
+HERE = os.path.dirname(__file__)
+ROOT = os.path.join(HERE, "..")
+
+try:
+    from mcp import ClientSession, StdioServerParameters
+    from mcp.client.stdio import stdio_client
+except ImportError:
+    ClientSession = None
+
+
+async def _run():
+    params = StdioServerParameters(command=sys.executable, args=[os.path.join(ROOT, "pdf_table_pull_lite_mcp.py")], cwd=ROOT)
+    async with stdio_client(params) as (r, w):
+        async with ClientSession(r, w) as s:
+            await s.initialize()
+            tools = [t.name for t in (await s.list_tools()).tools]
+            res = await s.call_tool("extract_pdf_tables", {"path": os.path.abspath(os.path.join(HERE, "fixtures", "ruled.pdf")), "pages": "1"})
+            return tools, res
+
+
+@unittest.skipIf(ClientSession is None, "mcp not installed")
+class MCP(unittest.TestCase):
+    def test_lists_and_calls_tool(self):
+        tools, res = asyncio.run(_run())
+        self.assertEqual(tools, ["extract_pdf_tables"])
+        self.assertFalse(res.isError)
+        text = " ".join(c.text for c in res.content if getattr(c, "type", "") == "text")
+        self.assertIn("Region,Q1,Q2", text)
+        self.assertIn("South,980,(40)", text)
+
+
+if __name__ == "__main__":
+    unittest.main()