Created
April 8, 2014 01:52
-
-
Save timvisher/10082108 to your computer and use it in GitHub Desktop.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
(ns sicp-scraper.core | |
(:require [net.cgrand.enlive-html :as html])) | |
(def ^:dynamic *html-resources* | |
(map (fn [page-number] | |
(-> (str "/Users/tim/Dropbox/sicp/full-text/mitpress.mit.edu/sicp/full-text/book/book-Z-H-" page-number ".html") | |
clojure.java.io/file | |
html/html-resource)) | |
(range 9 36))) | |
(defn node->anchor-title-tuple [node] | |
((juxt (fn [node] (last (.split (last (.split (get-in node [:attrs :href]) "#")) "%_toc_"))) html/text) node)) | |
(def ^:dynamic *chapter-frag->title* | |
(let [toc-resource (-> "/Users/tim/Dropbox/sicp/full-text/mitpress.mit.edu/sicp/full-text/book/book-Z-H-4.html" | |
clojure.java.io/file | |
html/html-resource)] | |
(into {} (map node->anchor-title-tuple | |
(html/select toc-resource [[:a | |
(complement (html/attr-contains :href "Temp")) | |
(html/attr-contains :href "chap")]]))))) | |
(def ^:dynamic *exercises-resource* | |
(html/html-resource (clojure.java.io/file (str "/Users/tim/Dropbox/sicp/full-text/mitpress.mit.edu/sicp/full-text/book/book-Z-H-37.html")))) | |
(defn interesting-exercise-nodes [] | |
(html/select *exercises-resource* #{[[:a | |
(html/attr? :href) | |
(html/attr-contains :href "thm")]]})) | |
(def ^:dynamic *exercise-anchor->title* | |
(let [tmp (into {} (map node->anchor-title-tuple (interesting-exercise-nodes)))] | |
(zipmap (keys tmp) (map (fn [val] (str "Exercise " val)) (vals tmp))))) | |
(def ^:dynamic *master-paged-order-nodes* | |
(map (fn [html-resource] | |
(html/select html-resource #{[#{:h1 :h2 :h3 :h4} | |
[:a | |
(html/attr? :href) | |
(complement (html/attr-contains :href "Temp")) | |
#{(html/attr-contains :href "sec") | |
(html/attr-contains :href "chap")}]] | |
[[:a | |
(html/attr? :name) | |
(html/pred #((apply hash-set (keys *exercise-anchor->title*)) (get-in % [:attrs :name])))]] | |
[[:a | |
(html/has [(html/text-pred #(re-seq #"Chapter" %))])]]})) | |
*html-resources*)) | |
(def ^:dynamic *section-anchors->headers* | |
(into {} | |
(map node->anchor-title-tuple | |
(reduce into [] (map (fn [order-node-page] | |
(filter (fn [node] | |
(get-in node [:attrs :href])) | |
order-node-page)) | |
*master-paged-order-nodes*))))) | |
(def ^:dynamic *master-order* | |
(reduce into | |
[] | |
(map (fn [order-node-page] | |
(map (fn [node] | |
(last (.split (or (get-in node [:attrs :name]) (get-in node [:attrs :href])) "%_toc_"))) | |
order-node-page)) | |
*master-paged-order-nodes*))) | |
(defn bare-anchor-fragment [anchor-node] | |
(last (.split (or (get-in anchor-node [:attrs :href]) | |
(get-in anchor-node [:attrs :name])) | |
"%_toc_"))) | |
(defn anchor-node->link [page-number anchor-node] | |
(let [page-base-link (str "http://mitpress.mit.edu/sicp/full-text/book/book-Z-H-" page-number ".html") | |
anchor-fragment (bare-anchor-fragment anchor-node)] | |
(str page-base-link "#" anchor-fragment))) | |
(defn link->fragment-link-tuple [link] | |
[(last (.split link "#")) link]) | |
(def *fragment->link* | |
(into {} | |
(map link->fragment-link-tuple | |
(reduce into [] | |
(map (fn [order-node-page page-number] | |
(map (partial anchor-node->link page-number) order-node-page)) | |
*master-paged-order-nodes* | |
(drop 9 (range))))))) | |
(defn title->org-header-marks [title] | |
(if (re-seq #"Exercise [0-9]+\.[0-9]+" title) | |
"***" | |
(apply str (repeat (+ 1 (count (re-seq #"\." title))) "*")))) | |
(def *org* | |
(map (fn [fragment] | |
(let [title (clojure.string/replace (or (*chapter-frag->title* fragment) | |
(*section-anchors->headers* fragment) | |
(*exercise-anchor->title* fragment)) | |
#" +" | |
" ")] | |
(format "%s TODO [[%s][%s]]" | |
(title->org-header-marks title) | |
(*fragment->link* fragment) | |
title))) | |
*master-order*)) | |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment