scrape-smith

No-dependency Python utilities for scraping workflows.

The package provides a scrape command with tools for table extraction, list extraction, body content JSONL conversion, combined three-in-one extraction, and polite file downloads.

HTML input is decoded from byte streams with charset hints from byte-order marks, HTTP headers, and HTML metadata, with UTF-8 and common Japanese encodings as fallbacks. This keeps non-ASCII page text readable across the extraction commands.

Installation

scrape-smith requires Python 3.10 or newer.

python -m pip install scrape-smith

Quickstart

scrape tables <html-file-or-url>

scrape tables page.html
scrape tables https://example.com/page.html -o result.csv
scrape tables page.html --format json -o tables.json

Without -o, output is written to a safe source-based filename:

page.html -> page-tables.csv
https://yeiichi.github.io/claim-class-model -> claim-class-model-tables.csv

If the default filename already exists, scrape avoids overwriting it by adding a numeric suffix, such as page-tables-2.csv.

On success, scrape reports the written file to stdout:

Wrote 1 table to page-tables.csv

Use --quiet to suppress the success report.

Extract lists from an HTML page:

scrape lists page.html
scrape lists https://example.com/page.html

The lists command reads a local HTML file or HTTP(S) URL and extracts ul, ol, and dl elements. Ordered and unordered lists produce a single item column in CSV. Definition lists produce two columns, term and description. Multiple lists are separated by blank rows.

Without -o, CSV output is written to a safe source-based filename:

page.html -> page-lists.csv

Convert body content from one HTML page to JSONL:

scrape content page.html
scrape content https://example.com/page.html

The content command reads a local HTML file or HTTP(S) URL, extracts visible text from the <body>, ignores metadata-like elements such as <script>, and writes one JSON object per line for headings (h1–h6), paragraphs (p), links (a), list items (li), table cells (td, th), block quotes (blockquote), figure captions (figcaption), definition terms and descriptions (dt, dd), table captions (caption), labels (label), and buttons (button):

{"h1": "Title"}
{"p": "Hello docs."}
{"a": "docs", "href": "/docs"}

Without -o, JSONL output is written to a safe source-based filename:

page.html -> page-content.jsonl

Extract all three in source order into one JSONL file:

scrape three page.html
scrape three https://example.com/page.html

The three command runs content, table, and list extraction in a single pass and writes every record to one JSONL file in document order. Each line is a {"type": ..., "data": ...} object where data is a JSON string:

{"type": "content", "data": "{\"tag\": \"h1\", \"text\": \"Title\"}"}
{"type": "list",    "data": "{\"tag\": \"ul\", \"items\": [\"Alpha\", \"Beta\"]}"}
{"type": "table",   "data": "{\"caption\": null, \"headers\": [\"Name\"], \"rows\": [[\"Ada\"]]}"}

Without -o, output is written to a safe source-based filename:

page.html -> page-extract.jsonl

Download target files from a URL list:

scrape download /path/to/urls.txt

The URL list should contain one URL per line. Blank lines and lines starting with # are ignored. scrape-smith downloads CSV, PDF, DOCX, XLSX, and PPTX files sequentially, waits between requests by default, and skips non-target URLs.

Without -o, files are saved under a collision-safe directory based on the URL list filename:

urls.txt -> urls-downloads/

Downloaded files keep their original filenames when the URL or response headers provide one, including UTF-8 response filenames and common percent-encoded non-ASCII URL paths. If no filename is available, scrape-smith falls back to a fixed-width epoch filename such as 1700000000123.pdf.

Warning: treat downloaded files as untrusted. Scan them before opening, and do not open files blindly; documents and spreadsheets can contain harmful content.

Command Line

The command-line entry point is scrape.

scrape tables <html-file-or-url> [--format csv|json] [-o OUTPUT] [--index N] [--quiet]
scrape lists <html-file-or-url> [--format csv|json] [-o OUTPUT] [--quiet]
scrape content <html-file-or-url> [-o OUTPUT] [--quiet]
scrape three <html-file-or-url> [-o OUTPUT] [--quiet]
scrape download <url-list-path> [-o OUTPUT_DIR] [--delay SECONDS]

Options:

--format: output format, either csv or json; defaults to csv
-o, --output: output file path; defaults to a safe source-based filename
--index: write one table by zero-based index
-q, --quiet: suppress success reports on stdout

Download options:

-o, --output-dir: output directory; defaults to a safe URL-list-based name
--delay: seconds to wait between requests; defaults to 1

Errors and validation messages are written to stderr. If --index is out of range, the command exits with status code 2.

Python API

Python APIs live under scrape_smith.tools.

from scrape_smith.tools.tables import extract_tables

tables = extract_tables("page.html")

Each extracted table has caption, headers, and rows fields. JSON output uses the same shape.

from scrape_smith.tools.downloads import download_files

summary = download_files("urls.txt", delay_seconds=1)
print(summary.downloaded_count)

from scrape_smith.tools.lists import extract_lists

lists = extract_lists("page.html")

from scrape_smith.tools.extract import extract_all

records = extract_all("page.html")

from scrape_smith.tools.content import extract_content_records

records = extract_content_records("page.html")

Documentation

Full documentation is available at https://scrape-smith.readthedocs.io/.

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

scrape_smith-0.3.1.tar.gz (13.3 kB view details)

Uploaded Jul 11, 2026 Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

The dropdown lists show the available interpreters, ABIs, and platforms. Enable javascript to be able to filter the list of wheel files.

scrape_smith-0.3.1-py3-none-any.whl (18.6 kB view details)

Uploaded Jul 11, 2026 Python 3

File details

Details for the file scrape_smith-0.3.1.tar.gz.

File metadata

Download URL: scrape_smith-0.3.1.tar.gz
Upload date: Jul 11, 2026
Size: 13.3 kB
Tags: Source
Uploaded using Trusted Publishing? Yes
Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for scrape_smith-0.3.1.tar.gz
Algorithm	Hash digest
SHA256	`2fd7b74f1ae85d503a25f887c58e769eb67e91e1d66d421e7dde1f10de905014`
MD5	`be6439f44b69b0bfb92b16a7d5943ed8`
BLAKE2b-256	`920e045e1383fc8995d021e6e4c6565c056acc13fe5f9de54e2b229e4d4f0f44`

See more details on using hashes here.

Provenance

The following attestation bundles were made for scrape_smith-0.3.1.tar.gz:

Publisher: publish-pypi.yml on yeiichi/scrape-smith

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Statement:
- Statement type: https://in-toto.io/Statement/v1
- Predicate type: https://docs.pypi.org/attestations/publish/v1
- Subject name: scrape_smith-0.3.1.tar.gz
- Subject digest: 2fd7b74f1ae85d503a25f887c58e769eb67e91e1d66d421e7dde1f10de905014
- Sigstore transparency entry: 2147496584
- Sigstore integration time: Jul 11, 2026
Source repository:
- Permalink: yeiichi/scrape-smith@5523d532de0efa3cf9cfc646df4e964e55c0f12c
- Branch / Tag: refs/heads/main
- Owner: https://github.com/yeiichi
- Access: public
Publication detail:
- Token Issuer: https://token.actions.githubusercontent.com
- Runner Environment: github-hosted
- Publication workflow: publish-pypi.yml@5523d532de0efa3cf9cfc646df4e964e55c0f12c
- Trigger Event: push