qry 0.2.12

Ultra-fast file search and processing tool

qry

Ultra-fast file search and metadata extraction tool.

Features

• Fast filesystem search with optional depth, date, and size filters
• Optimized performance — caching, parallel processing, date-based directory pruning
• CLI modes: search, interactive, batch, version
• HTTP API (FastAPI) for JSON and HTML search responses
• Metadata extraction for matched files (size, timestamps, content type)
• Streaming results — Ctrl+C stops search mid-way and outputs what was found so far
• Smart directory exclusions — .git, .venv, __pycache__, dist, node_modules skipped by default
• YAML, JSON, and paths output — machine-readable output for piping into other tools
• Python API — import qry; qry.search(...) for use in other applications
• Regex search — --regex flag for pattern matching in filenames and content
• Size filtering — --min-size / --max-size with human-readable units (1k, 10MB, 1G)
• Sort results — --sort name|size|date
• Content preview — --preview shows matching line with context for content search
• Date filtering — --last-days, --after-date, --before-date for time-based searches
• Parallel search — -w/--workers for multi-threaded directory processing
• Priority-based search — searches important directories first (src/, tests/), then lower priority (cache/, .git/)
• Incremental results — shows results as they're found with timeout-based priority fallback

Installation

Poetry (recommended)

poetry install --with dev

pip (minimal)

pip install -r requirements.txt

Quick start

# search in current directory (filename match, YAML output)
poetry run qry "invoice"

# search file contents with preview snippet
poetry run qry "def search" -c -P --path ./qry

# regex search, sorted by name
poetry run qry "\.py$" -r --sort name --path .

# pipe-friendly output for shell pipelines
poetry run qry "TODO" -c -o paths | xargs grep -n "FIXME"

# show version and engines
poetry run qry version

CLI usage

qry search [query ...] [-f] [-c] [-r] [-P] [--type EXT1,EXT2] [--scope PATH | --path PATH]
           [--depth N] [--last-days N] [--limit N] [--min-size SIZE] [--max-size SIZE]
           [--sort name|size|date] [--exclude DIR] [--no-exclude]
           [--output yaml|json|paths]

qry interactive
qry batch <input_file> [--output-file FILE] [--format text|json|csv] [--workers N]
qry version

Search mode flags

Flag	Long form	Searches
(none)		filename (default)
-f	--filename	filename only
-c	--content	file contents
-r	--regex	treat query as regular expression

Filtering flags

Flag	Description
-t EXT	Filter by file type (comma-separated)
-d N	Max directory depth
-l N	Limit results (0 = unlimited, default)
--last-days N	Files modified in last N days
--after-date YYYY-MM-DD	Files modified after date
--before-date YYYY-MM-DD	Files modified before date
-w N	Workers for parallel search (default: 4)
--min-size SIZE	Minimum file size (e.g. 1k, 10MB, 1G)
--max-size SIZE	Maximum file size (e.g. 100k, 5MB)
-e DIR	Exclude extra directory (repeatable, comma-separated)
--no-exclude	Disable all default exclusions

Output flags

Flag	Description
-o yaml	YAML output (default)
-o json	JSON output
-o paths	One path per line — pipe-friendly
-P	--preview — show matching line with context (with -c)
--sort	Sort results by name, size, or date

Default excluded directories: .git .venv __pycache__ dist node_modules .tox .mypy_cache

Priority-based search

When priority mode is enabled, directories are searched in order of importance:

Priority	Value	Directories
SOURCE	100	src/, source/, lib/, code/
PROJECT	90	tests/, test/, docs/, scripts/, examples/
CONFIG	80	config/, .config/, settings/
MAIN	70	main/, app/, core/, server/, client/
MODULES	60	modules/, module/, components/, packages/, plugins/
UTILS	50	utils/, helpers/, tools/
BUILD	40	build/, dist/, out/, target/, release/
CACHE	30	cache/, __pycache__/, node_modules/, .pytest_cache/
TEMP	20	temp/, tmp/, .tmp/
GENERATED	10	generated/, compiled/, bin/, obj/
EXCLUDED	0	.git/, .svn/, .venv/, venv/, .idea/, .vscode/

This ensures that important directories (source code) are searched first, while cache and temporary directories are searched last.

Examples

# search by filename (default)
poetry run qry "invoice"

# search inside file contents — press Ctrl+C to stop early
poetry run qry "def search" -c
poetry run qry "TODO OR FIXME" -c --type py --path ./src

# regex search for Python files
poetry run qry "\.py$" -r --sort name -s qry/

# content search with preview snippet
poetry run qry "search" -c -P --sort name -s qry/ -d 2

# filter by file size
poetry run qry "" --min-size 10k --max-size 1MB --sort size

# JSON output for piping
poetry run qry "invoice" -o json | jq '.results[]'

# pipe-friendly: one path per line
poetry run qry "TODO" -c -o paths | xargs grep -n "FIXME"
poetry run qry "invoice" -o paths | xargs -I{} cp {} /backup/

# exclude extra directories
poetry run qry "config" -e build -e ".cache"

# disable all exclusions (search everything)
poetry run qry "config" --no-exclude

# combine scope/depth/type/date
poetry run qry "invoice OR faktura" --scope /data/docs --depth 3
poetry run qry search "report" --type pdf,docx --last-days 7
poetry run qry batch queries.txt --format json --output-file results.json

# date filtering examples
poetry run qry "report" --last-days 30          # files modified in last 30 days
poetry run qry "invoice" --after-date 2026-01-01    # files after Jan 1, 2026
poetry run qry "invoice" --before-date 2025-12-31  # files before Dec 31, 2025
poetry run qry "report" --after-date 2026-01-01 --before-date 2026-02-01  # date range

Python API

Use qry directly from Python — no subprocess needed:

import qry

# Return all matching file paths as a list
files = qry.search("invoice", scope="/data/docs", mode="content", depth=3)

# Stream results one at a time (memory-efficient, supports Ctrl+C)
for path in qry.search_iter("TODO", scope="./src", mode="content"):
    print(path)

# Regex search with sorting
py_files = qry.search(r"test_.*\.py$", scope=".", regex=True, sort_by="name")

# Size filtering — find large files
big = qry.search("", scope=".", min_size=1024*1024, sort_by="size")

# Custom exclusions
files = qry.search("config", exclude_dirs=[".git", "build", ".venv"])

Parameters for both qry.search() and qry.search_iter():

Parameter	Type	Default	Description
query_text	str	—	Text to search for
scope	str	"."	Directory to search
mode	str	"filename"	"filename", "content", or "both"
depth	int|None	None	Max directory depth
file_types	list|None	None	Extensions to include, e.g. ["py","txt"]
exclude_dirs	list|None	None	Dir names to skip (None = use defaults)
max_results	int	unlimited	Hard cap on results
min_size	int|None	None	Minimum file size in bytes
max_size	int|None	None	Maximum file size in bytes
regex	bool	False	Treat query as regular expression
sort_by	str|None	None	Sort by "name", "size", or "date"
date_range	tuple|None	None	Date range as (start_date, end_date) datetime tuples

Example with date filtering:

from datetime import datetime, timedelta
import qry

# Files modified in last 30 days
end_date = datetime.now()
start_date = end_date - timedelta(days=30)
files = qry.search("invoice", date_range=(start_date, end_date))

# Files modified in 2026
files = qry.search("report", date_range=(datetime(2026, 1, 1), datetime(2026, 12, 31)))

HTTP API usage

Run server:

poetry run qry-api --host 127.0.0.1 --port 8000

Main endpoints:

• GET /api/search
• GET /api/search/html
• GET /api/engines
• GET /api/health
• OpenAPI docs: GET /api/docs

Development

Run tests

poetry run pytest -q

Useful make targets

make install
make test
make lint
make type-check
make run-api

Project structure

• qry/cli/ – CLI commands and interactive mode
• qry/api/ – FastAPI application and routes
• qry/core/ – core data models
• qry/engines/ – search engine implementations
• qry/web/ – HTML renderer/templates integration
• tests/ – test suite

Additional docs

• Usage examples: EXAMPLES.md

License

Apache License 2.0 - see LICENSE for details.

Author

Created by Tom Sapletta - [e-mail]