liteparse

Name: liteparse
Author: K-Dense-AI

433installs3.1kforks

How to install liteparse?

k-dense-ai/scientific-agent-skills/liteparse

$npx -y skills add K-Dense-AI/scientific-agent-skills --skill liteparse

Installs into the current project.

Summary

Local document and PDF parsing with spatial text and bounding boxes. Use for extracting text from PDFs, DOCX, Office files, and images; OCR on scans; layout-preserved JSON for RAG; batch-ingesting paper folders; or page screenshots for multimodal agents — even when the user does

Files · 1

View on GitHub

SKILL.md

1	# LiteParse — Local Document Parsing
2
3	## Overview
4
5	LiteParse is a fast, open-source document parser (Rust core, Python/Node bindings) focused on local, layout-aware text extraction with bounding boxes. It does not produce Markdown and does not call cloud LLMs. Outputs are plain text (layout-preserved) or structured JSON with per-page `text_items` (position, font metadata, optional confidence).
6
7	Version note: Examples target liteparse 2.0.0 (PyPI, May 2026). The upstream V1 branch is legacy; this skill documents V2 / main only.
8
9	For parser selection vs MarkItDown, the `pdf` skill, or LlamaParse, see `references/choosing_a_parser.md`.
10
11	## When to Use This Skill
12
13	Use LiteParse when you need:
14
15	- Fast local parsing of PDFs or converted Office/image files without cloud dependencies
16	- Spatial text with bounding boxes for layout-aware RAG, citation grounding, or figure/table region logic
17	- OCR on scanned PDFs or images (bundled Tesseract, or a user-run HTTP OCR server)
18	- Page screenshots (PNG) for multimodal agents that must see charts, figures, or handwriting
19	- Batch ingestion of literature folders, supplementary PDFs, or protocol libraries
20	- Page subsets or password-protected PDFs
21
22	## When Not to Use
23
24	\| Task \| Use instead \|
25	\|------\|-------------\|
26	\| Markdown for LLM ingestion (EPUB, audio, YouTube, HTML) \| `markitdown` skill \|
27	\| Merge/split PDFs, forms, watermarks, rotation \| `pdf` skill \|
28	\| Dense tables, handwriting, production cloud pipelines \| [LlamaParse](https://docs.cloud.llamaindex.ai/llamaparse/overview) (cloud; sign up separately) \|
29
30	## Installation
31
32	```bash
33	uv pip install "liteparse==2.0.0"
34	```
35
36	This installs the Python bindings and the `lit` CLI. Verify:
37
38	```bash
39	lit --help
40	python -c "import liteparse; print(liteparse.__version__)"
41	```
42
43	Optional system tools (for non-PDF inputs):
44
45	- LibreOffice — Word, Excel, PowerPoint, OpenDocument, CSV/TSV
46	- ImageMagick — PNG, JPEG, TIFF, WebP, SVG, etc.
47
48	Install commands are in `references/ocr_and_formats.md`.
49
50	Node.js / TypeScript (optional): `npm i @llamaindex/liteparse` — see `references/api_reference.md`.
51
52	---
53
54	## Quick Start
55
56	### Python
57
58	```python
59	from liteparse import LiteParse
60
61	parser = LiteParse(quiet=True)
62	result = parser.parse("paper.pdf")
63	print(result.text)
64
65	for page in result.pages:
66	print(f"Page {page.page_num}: {len(page.text_items)} items")
67	```
68
69	### CLI
70
71	```bash
72	# Layout-preserved text (default)
73	lit parse paper.pdf
74
75	# Structured JSON with bounding boxes
76	lit parse paper.pdf --format json -o paper.json
77
78	# Disable OCR on text-native PDFs (faster)
79	lit parse paper.pdf --no-ocr
80	```
81
82	---
83
84	## Core Workflows
85
86	### 1. Parse to layout-preserved text
87
88	Best for quick full-document text or feeding chunkers that do not need coordinates.
89
90	```python
91	parser = LiteParse(ocr_enabled=True, quiet=True)
92	result = parser.parse("document.pdf")
93	full_text = result.text
94	```
95
96	```bash
97	lit parse document.pdf -o output.txt
98	```
99
100	### 2. Parse to structured JSON (bounding boxes)
101
102	Use when building layout-aware RAG, highlighting source regions, or joining text with screenshots.
103
104	```python
105	import json
106	from liteparse import LiteParse
107
108	parser = LiteParse(output_format="json", quiet=True)
109	result = parser.parse("document.pdf")
110
111	# Programmatic access
112	for page in result.pages:
113	for item in page.text_items:
114	bbox = (item.x, item.y, item.width, item.height)
115	# item.text, item.confidence, item.font_name, item.font_size
116	```
117
118	```bash
119	lit parse document.pdf --format json -o document.json
120	```
121
122	JSON field layout: `references/output_formats.md`.
123
124	### 3. Parse specific pages
125
126	```python
127	parser = LiteParse(target_pages="1-5,10,15-20", quiet=True)
128	result = parser.parse("long_paper.pdf")
129	```
130
131	```bash
132	lit parse long_paper.pdf --target-pages "1-5,10"
133	```
134
135	### 4. Parse from bytes or stdin
136
137	Useful for uploads, S3 downloads, or piping remote PDFs.
138
139	```python
140	with open("document.pdf", "rb") as f:
141	result = parser.parse(f.read())
142	```
143
144	```bash
145	curl -sL https://example.com/report.pdf \| lit parse -
146	```
147
148	### 5. Page screenshots for multimodal agents
149
150	Screenshots capture visual content that text extraction alone misses (figures, complex tables, handwriting).
151
152	```python
153	from pathlib i

People who liked this also viewed

6 picks

Type

liteparse

433installs3.1kforks

View on GitHub

How to install liteparse?

k-dense-ai/scientific-agent-skills/liteparse

$npx -y skills add K-Dense-AI/scientific-agent-skills --skill liteparse

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# LiteParse — Local Document Parsing
2
3	## Overview
4
5	LiteParse is a fast, open-source document parser (Rust core, Python/Node bindings) focused on local, layout-aware text extraction with bounding boxes. It does not produce Markdown and does not call cloud LLMs. Outputs are plain text (layout-preserved) or structured JSON with per-page `text_items` (position, font metadata, optional confidence).
6
7	Version note: Examples target liteparse 2.0.0 (PyPI, May 2026). The upstream V1 branch is legacy; this skill documents V2 / main only.
8
9	For parser selection vs MarkItDown, the `pdf` skill, or LlamaParse, see `references/choosing_a_parser.md`.
10
11	## When to Use This Skill
12
13	Use LiteParse when you need:
14
15	- Fast local parsing of PDFs or converted Office/image files without cloud dependencies
16	- Spatial text with bounding boxes for layout-aware RAG, citation grounding, or figure/table region logic
17	- OCR on scanned PDFs or images (bundled Tesseract, or a user-run HTTP OCR server)
18	- Page screenshots (PNG) for multimodal agents that must see charts, figures, or handwriting
19	- Batch ingestion of literature folders, supplementary PDFs, or protocol libraries
20	- Page subsets or password-protected PDFs
21
22	## When Not to Use
23
24	\| Task \| Use instead \|
25	\|------\|-------------\|
26	\| Markdown for LLM ingestion (EPUB, audio, YouTube, HTML) \| `markitdown` skill \|
27	\| Merge/split PDFs, forms, watermarks, rotation \| `pdf` skill \|
28	\| Dense tables, handwriting, production cloud pipelines \| [LlamaParse](https://docs.cloud.llamaindex.ai/llamaparse/overview) (cloud; sign up separately) \|
29
30	## Installation
31
32	```bash
33	uv pip install "liteparse==2.0.0"
34	```
35
36	This installs the Python bindings and the `lit` CLI. Verify:
37
38	```bash
39	lit --help
40	python -c "import liteparse; print(liteparse.__version__)"
41	```
42
43	Optional system tools (for non-PDF inputs):
44
45	- LibreOffice — Word, Excel, PowerPoint, OpenDocument, CSV/TSV
46	- ImageMagick — PNG, JPEG, TIFF, WebP, SVG, etc.
47
48	Install commands are in `references/ocr_and_formats.md`.
49
50	Node.js / TypeScript (optional): `npm i @llamaindex/liteparse` — see `references/api_reference.md`.
51
52	---
53
54	## Quick Start
55
56	### Python
57
58	```python
59	from liteparse import LiteParse
60
61	parser = LiteParse(quiet=True)
62	result = parser.parse("paper.pdf")
63	print(result.text)
64
65	for page in result.pages:
66	print(f"Page {page.page_num}: {len(page.text_items)} items")
67	```
68
69	### CLI
70
71	```bash
72	# Layout-preserved text (default)
73	lit parse paper.pdf
74
75	# Structured JSON with bounding boxes
76	lit parse paper.pdf --format json -o paper.json
77
78	# Disable OCR on text-native PDFs (faster)
79	lit parse paper.pdf --no-ocr
80	```
81
82	---
83
84	## Core Workflows
85
86	### 1. Parse to layout-preserved text
87
88	Best for quick full-document text or feeding chunkers that do not need coordinates.
89
90	```python
91	parser = LiteParse(ocr_enabled=True, quiet=True)
92	result = parser.parse("document.pdf")
93	full_text = result.text
94	```
95
96	```bash
97	lit parse document.pdf -o output.txt
98	```
99
100	### 2. Parse to structured JSON (bounding boxes)
101
102	Use when building layout-aware RAG, highlighting source regions, or joining text with screenshots.
103
104	```python
105	import json
106	from liteparse import LiteParse
107
108	parser = LiteParse(output_format="json", quiet=True)
109	result = parser.parse("document.pdf")
110
111	# Programmatic access
112	for page in result.pages:
113	for item in page.text_items:
114	bbox = (item.x, item.y, item.width, item.height)
115	# item.text, item.confidence, item.font_name, item.font_size
116	```
117
118	```bash
119	lit parse document.pdf --format json -o document.json
120	```
121
122	JSON field layout: `references/output_formats.md`.
123
124	### 3. Parse specific pages
125
126	```python
127	parser = LiteParse(target_pages="1-5,10,15-20", quiet=True)
128	result = parser.parse("long_paper.pdf")
129	```
130
131	```bash
132	lit parse long_paper.pdf --target-pages "1-5,10"
133	```
134
135	### 4. Parse from bytes or stdin
136
137	Useful for uploads, S3 downloads, or piping remote PDFs.
138
139	```python
140	with open("document.pdf", "rb") as f:
141	result = parser.parse(f.read())
142	```
143
144	```bash
145	curl -sL https://example.com/report.pdf \| lit parse -
146	```
147
148	### 5. Page screenshots for multimodal agents
149
150	Screenshots capture visual content that text extraction alone misses (figures, complex tables, handwriting).
151
152	```python
153	from pathlib i

People who liked this also viewed

6 picks

Type