{ "slug": "rule-based-extractor", "name": "gizza-ai/rule-based-extractor", "version": "0.1.0", "title": "Rule-Based Text Extractor — Regex Rules to JSON or CSV — gizza.ai", "description": "Turn logs, invoices or scraped text into structured data with named regex rules. Grok-style %{PATTERN:field} shortcuts, JSON/CSV output, per-rule report.", "tags": [ "rule based extractor", "regex to json", "grok pattern parser", "log parser online", "extract fields from text", "text to csv converter", "named capture groups", "structured data extraction" ], "category": "text", "urls": { "page": "https://gizza.ai/tools/rule-based-extractor/", "markdown": "https://gizza.ai/tools/rule-based-extractor/index.md", "descriptor": "https://gizza.ai/tools/rule-based-extractor/tool.json", "deep_link_example": "https://gizza.ai/tools/rule-based-extractor/?text=Paste%20the%20log%20lines%2C%20invoice%20text%20or%20records%20to%20extract%20from%E2%80%A6&rules=date%20%3D%20%25%7BDATE_ISO%7D%0Alevel%20%3D%20%25%7BLOGLEVEL%7D%0Aclient%20%3D%20%25%7BIPV4%7D&split=whole&split_pattern=%5E-%7B3%2C%7D%24&matches=first&ignore_case=true&multiline=true&dotall=true&trim=true&unique=true&on_missing=skip&skip_empty_records=true&max_records=5000&max_matches=1000&output=json&pretty=true" }, "cli": "gizza tool rule-based-extractor \"Paste the log lines, invoice text or records to extract from…\" 'rules=date = %{DATE_ISO}\nlevel = %{LOGLEVEL}\nclient = %{IPV4}'", "tool": { "description": "Apply a set of named regex rules to unstructured text and return the captured fields as JSON, CSV, a readable listing, or a rule-by-rule report. Write one rule per line as 'field = regex'; a bare regex line contributes every named group it contains. Grok-style placeholders keep rules short — %{DATE_ISO:date}, %{IPV4:client}, %{LOGLEVEL:level} and ~30 other built-ins — and '@NAME = regex' defines your own. Set split to lines/paragraphs/pattern to run the rules per record and get one object per record, matches='all' to collect every hit as an array, and output='report' to see which rules never matched. Good for turning log lines, invoices, emails or scraped text into structured data.", "parameters": { "additionalProperties": false, "properties": { "dotall": { "default": false, "description": "Let '.' match newlines too (the regex 's' flag), so a rule can span several lines of a record.", "type": "boolean" }, "ignore_case": { "default": false, "description": "Match rules case-insensitively (the regex 'i' flag). Per-rule inline flags like (?i) also work.", "type": "boolean" }, "matches": { "default": "first", "description": "Whether each rule keeps only its first match per record (default) or every match. 'all' makes each field a JSON array.", "enum": [ "first", "all" ], "type": "string" }, "max_matches": { "default": 1000, "description": "Safety cap on how many times one rule may match inside a single record when matches='all'. Exceeding it is an error.", "maximum": 10000, "minimum": 1, "type": "integer" }, "max_records": { "default": 5000, "description": "Safety cap on how many records the input may split into. Exceeding it is an error, never a silent truncation.", "maximum": 50000, "minimum": 1, "type": "integer" }, "multiline": { "default": false, "description": "Make ^ and $ match at every line break inside a record instead of only at its start and end (the regex 'm' flag).", "type": "boolean" }, "on_missing": { "default": "skip", "description": "What to do when a rule matches nothing in a record: 'skip' (default) omits the key, 'null' emits it as null/empty so every record has the same shape, 'error' fails with the field name.", "enum": [ "skip", "null", "error" ], "type": "string" }, "output": { "default": "json", "description": "Result format: 'json' (default) objects of named fields, 'csv' one row per record, 'text' a readable field listing, or 'report' a per-rule hit count that shows which rules never matched.", "enum": [ "json", "csv", "text", "report" ], "type": "string" }, "pretty": { "default": false, "description": "Indent the JSON output. Applies to output='json' only.", "type": "boolean" }, "rules": { "description": "One rule per line. 'field = regex' captures one named field; a bare regex line captures every named group in it. '@NAME = regex' defines a reusable pattern, and '%{NAME}' / '%{NAME:field}' inserts a built-in or user pattern (WORD, INT, NUMBER, EMAIL, URL, IPV4, IPV6, MAC, UUID, HASH, DATE_ISO, DATE_US, TIME, TIMESTAMP_ISO, SYSLOG_TIME, YEAR, LOGLEVEL, HTTP_METHOD, HTTP_STATUS, HOSTNAME, PATH, QUOTED, MONEY, PERCENT, PHONE, ZIP_US, SEMVER, TICKET, GREEDYDATA, DATA, HEX, NOTSPACE, SPACE, IP). Lines starting with # or // are comments. Max 200 rules.", "type": "string" }, "skip_empty_records": { "default": true, "description": "Drop records where no rule matched at all. On by default; ignored when split='whole'.", "type": "boolean" }, "split": { "default": "whole", "description": "How the text is divided into records before the rules run. 'whole' (default) treats the input as one record; 'lines' and 'paragraphs' produce one output object per line or blank-line-separated block; 'pattern' splits on the regex in split_pattern.", "enum": [ "whole", "lines", "paragraphs", "pattern" ], "type": "string" }, "split_pattern": { "default": "", "description": "Record-separator regex, used only when split='pattern'. For example '^-{3,}$' with multiline=true splits on a line of dashes.", "type": "string" }, "text": { "description": "The text to extract from. Up to 1 MB.", "type": "string" }, "trim": { "default": true, "description": "Strip leading and trailing whitespace from every captured value. On by default.", "type": "boolean" }, "unique": { "default": false, "description": "With matches='all', drop repeated values within a field, keeping first-seen order.", "type": "boolean" } }, "required": [ "text", "rules" ], "type": "object" } } }