codes.clj.docs/extractor

Tool to extract namespace/functions documentation from Clojure projects into indexed datalevin file.

Config resources/config.edn

This file has two main functions:

• :db key is defined where the datalevin files will be placed.
• :deps key is defined which libraries will be downloaded, parsed and indexed in the database
  • This configuration is made using the same clojure.tools.deps's git coordinates with some extra keys:
    • :project/group use to overide the project group, currently extracted from the lib ORG
    • :deps/manifest :deps since we use tools.deps we need to set this on pure lein projects to force the download of the lib.
    • :project/source-paths if a lein lib uses :source-paths, like reitit, we need to add this information here to indicate to clj-kondo where are the files to analyze.

CLI

Extract and generate datalevin file

clojure -X:extract

Flow

sequenceDiagram;
autonumber
participant C as Config
participant A as Analysis
participant AD as Adapters
participant DB as Datalevin
participant CI
participant RE as Release 

C->>A: Read file "resources/config.edn"
A->>A: Parse config projects to download
Note left of A: tools.deps
A->>A: Download projects
Note left of A: clj-kondo
A->>AD: Parse projects function/docs analysis
AD->>DB: Adapts & Indexes data into datoms
DB->>CI: Bulk-transact all datoms
CI->>RE: Zip and Publish

Loading

Using

Go to this github release page, download and unzip the docs-db.zip file.

Connecting

This database depends on datalevin/datalevin {:mvn/version "0.9.3"} onwards.

Since v.0.3.0, because of the new full-text index analyzers, this database requires this minimal connection opts to be used:

(require '[datalevin.core :as d]
         '[datalevin.search-utils :as su]
         '[datalevin.interpret :refer [inter-fn]])

(defn merge-tokenizers
  "Merges the results of tokenizer a and b into one sequence."
  [tokenizer-a tokenizer-b]
  (inter-fn [^String s]
    (into (sequence (tokenizer-a s))
      (sequence (tokenizer-b s)))))

(def conn-opts
  (let [query-analyzer (su/create-analyzer
                        {:tokenizer (merge-tokenizers
                                     (inter-fn [s] [[s 0 0]])
                                     (su/create-regexp-tokenizer #"[\s:/\.;,!=?\"'()\[\]{}|<>&@#^*\\~`\-]+"))
                         :token-filters [su/lower-case-token-filter]})]
    {:search-domains {"project-name" {:query-analyzer query-analyzer}
                      "namespace-name" {:query-analyzer query-analyzer}
                      "definition-name" {:query-analyzer query-analyzer}}}))

(d/get-conn "path/to/db" nil conn-opts)

Tools used

• tools.deps Download projects/libraries
• clj-kondo Parses/Analyses projects data
• datalevin Indexing and Storage of the data

Developing

Repl

clojure -M:dev:nrepl

Tests

clojure -M:dev:test

Build

clojure -T:build uberjar

Lint

clojure -M:clojure-lsp diagnostics
clojure -M:clojure-lsp clean-ns
clojure -M:clojure-lsp format

Other iterations

• https://github.com/rafaeldelboni/clojure-document-extractor

License

This is free and unencumbered software released into the public domain.
For more information, please refer to http://unlicense.org