bsqb 0.2.0

Type-safe query builder for Brave Search operators

bsqb

bsqb (Brave Search Query Builder) is a type-safe, zero-dependency Python library for constructing Brave Search query strings using the official search operators.

It helps you build valid q parameters for the Brave Search API with a fluent API, automatic quoting, and validation against API limits.

Features

• Complete operator coverage — All operators from the official Brave Search documentation
• Fluent API — Chain methods for readable, composable queries
• Type-safe — Full type hints and py.typed marker for mypy/Pylance
• Zero dependencies — Lightweight core with no runtime requirements
• API limit validation — Enforces the 400 character / 50 word limits on q
• Logical operators — AND, OR, NOT with Python operators (&, |, ~)

Installation

pip install bsqb

Quick start

from bsqb import Query

# Basic query with field operators
query = Query("machine learning").filetype("pdf").lang("en")
print(query.build())
# machine learning filetype:pdf lang:en

# Use with the Brave Search API
import urllib.parse
import urllib.request

params = urllib.parse.urlencode({"q": query.build()})
url = f"https://api.search.brave.com/res/v1/web/search?{params}"
request = urllib.request.Request(
    url,
    headers={"X-Subscription-Token": "YOUR_API_KEY"},
)

Supported operators

Operator	Method	Example output
Plain term	Query("term")	term
Exact phrase	.phrase("exact phrase")	"exact phrase"
Force include	.include("term")	+term
Exclude	.exclude("term")	-term
File extension	.ext("pdf")	ext:pdf
File type	.filetype("pdf")	filetype:pdf
In title	.intitle("2023")	intitle:2023
In body	.inbody("keyword")	inbody:keyword
In page	.inpage("keyword")	inpage:keyword
Language (ISO 639-1)	.lang("es")	lang:es
Language alias	.language("es")	language:es
Location (ISO 3166-1)	.loc("gb")	loc:gb
Location alias	.location("gb")	location:gb
Site / domain	.site("example.com")	site:example.com
Logical AND	.and_(other) or &	term1 AND term2
Logical OR	.or_(other) or |	term1 OR term2
Logical NOT	.not_(other) or ~	NOT term

Operators can be placed anywhere in the query string, matching Brave Search behavior.

Most operator methods accept a single value or a list/tuple of values:

Query("AI startup").exclude(["google", "microsoft", "amazon"])
Query("news").site(["reuters.com", "bloomberg.com"])

Examples

Official documentation examples

from bsqb import Query

# Academic research
Query("climate change").filetype("pdf").site("edu").intitle("2024").build()
# climate change filetype:pdf site:edu intitle:2024

# Multilingual content
Query("recettes cuisine").loc("ca").lang("fr").build()
# recettes cuisine loc:ca lang:fr

# Competitive analysis
Query("AI startup").exclude(["google", "microsoft", "amazon", "meta"]).build()
# AI startup -google -microsoft -amazon -meta

# Technical documentation
(
    Query("python")
    .phrase("asyncio")
    .intitle("documentation")
    .site("docs.python.org")
    .build()
)
# python "asyncio" intitle:documentation site:docs.python.org

Logical operators

from bsqb import Query, combine_and, combine_or

# AND — visa info in English from UK sites
Query("visa").loc("gb").and_(Query().lang("en")).build()
# visa loc:gb AND lang:en

# OR — travel requirements for Australia or New Zealand
(
    Query("travel requirements")
    .inpage("australia")
    .or_(Query().inpage("new zealand"))
    .build()
)
# travel requirements inpage:australia OR inpage:"new zealand"

# NOT — exclude a domain
Query("brave search").not_(Query().site("brave.com")).build()
# brave search NOT site:brave.com

# Python operators
(Query("coffee") | Query("tea")).exclude("starbucks").build()
# coffee OR tea -starbucks

# Combine multiple queries
combine_and(Query("visa").loc("gb"), Query().lang("en")).build()
combine_or(Query().site("reuters.com"), Query().site("bloomberg.com")).build()
Query("news").site(["reuters.com", "bloomberg.com"]).build()

Advanced usage

from bsqb import Query, phrase, raw, term

# Build from AST nodes
Query.from_nodes(term("python"), phrase("asyncio"), raw("site:docs.python.org"))

# Wrap an existing query string
Query.parse("machine learning filetype:pdf lang:en").build()

# Skip validation for edge cases
Query.parse("...").build(validate=False)

Validation

The Brave Search API enforces these limits on the q parameter:

• Maximum 400 characters
• Maximum 50 words
• Query cannot be empty

Call .build() to validate (default), or .render() / str() to get the string without validation:

from bsqb import Query, QueryValidationError, EmptyQueryError

query = Query("hello world")

query.render()   # "hello world" — no validation
query.build()    # "hello world" — validates limits

try:
    Query().build()
except EmptyQueryError:
    ...

try:
    Query.parse(" ".join(["word"] * 51)).build()
except QueryValidationError as exc:
    print(exc.query)

Integration with Brave Search API

Search operators are included in the q parameter. Set operators=true (the default) in API requests:

import urllib.parse
import urllib.request

from bsqb import Query

query = Query("python").phrase("asyncio").filetype("pdf").lang("en")

params = {
    "q": query.build(),
    "count": "10",
    "operators": "true",
}
url = "https://api.search.brave.com/res/v1/web/search?" + urllib.parse.urlencode(params)

request = urllib.request.Request(
    url,
    headers={
        "Accept": "application/json",
        "X-Subscription-Token": "YOUR_API_KEY",
    },
)

For POST requests with long queries, pass the built string as the q field in the request body.

Development

git clone https://github.com/kodzghly/bsqb.git
cd bsqb
python -m pip install -e ".[dev]"

# Run tests
pytest

# Lint and type check
ruff check src tests
mypy src

Publishing to PyPI

This package uses Hatchling as the build backend.

First-time setup

1. Create accounts on PyPI and TestPyPI
2. Enable 2FA on both accounts
3. Create an API token at pypi.org/manage/account/token/

Test on TestPyPI

python -m pip install --upgrade build twine

# Bump version in pyproject.toml and src/bsqb/__init__.py
python -m build

# Upload to TestPyPI
twine upload --repository testpypi dist/*

# Verify installation
pip install --index-url https://test.pypi.org/simple/ bsqb

Publish to PyPI

python -m build
twine check dist/*
twine upload dist/*

Recommended: publish via GitHub Actions

Add trusted publishing on PyPI for your repository, then create a release workflow triggered by git tags:

# .github/workflows/publish.yml
name: Publish

on:
  release:
    types: [published]

jobs:
  publish:
    runs-on: ubuntu-latest
    permissions:
      id-token: write
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-python@v5
        with:
          python-version: "3.12"
      - run: pip install build
      - run: python -m build
      - uses: pypa/gh-action-pypi-publish@release/v1

Create a GitHub release with tag v0.2.0 to trigger publication.

References

• Brave Search Operators (API docs)
• Brave Search Operators (help page)
• Brave Web Search API

Support

If bsqb is useful to you, consider sponsoring on GitHub — it helps keep the project maintained.

License

MIT — see LICENSE.