first commit

1 files changed, 111 insertions, 0 deletions

diff --git a/ankivoc.py b/ankivoc.py
new file mode 100644
index 0000000..da8adb2
--- /dev/null
+++ b/ankivoc.py
@@ -0,0 +1,111 @@
+#!/usr/bin/env python3
+
+from bs4 import BeautifulSoup, Comment
+import requests
+import sys
+import re
+
+def lookup(word):
+    r = requests.get("https://www.oxforddictionaries.com/definition/english/" +
+                     word)
+    soup = BeautifulSoup(r.text, "html.parser")
+
+    entry = "<meta charset=\"utf-8\">"
+
+    for res in soup.find_all("section", class_="senseGroup"):
+        entry += res.find_all("span", class_="partOfSpeech")[0].get_text()
+
+        # remove comments
+        for el in res(text=lambda text: isinstance(text, Comment)):
+            el.extract()
+
+        # remove annyoing link
+        for el in res.find_all("a", class_="moreInformationSynonyms"):
+            el.extract()
+
+        for sense in res.find_all("div", class_="msDict"):
+            # prepare example sentences for "more" tag
+            # The class typo is real.
+            for el in sense.find_all("a", class_="moreInformationExemples"):
+                #el.name = "button"  # a
+                el["href"] = "#"
+                el["onclick"] = \
+                    "var els = this.parentNode.getElementsByTagName('ul')[0].getElementsByTagName('li');" + \
+                    "for(var c = 0; c < els.length; c++){" + \
+                        "if (els[c].getAttribute('data-toggled') == 'false') {" + \
+                            "els[c].style = 'display: none;';" + \
+                            "els[c].setAttribute('data-toggled', 'true');" + \
+                        "} else {" + \
+                            "els[c].style = '';" + \
+                            "els[c].setAttribute('data-toggled', 'false');" + \
+                        "}" + \
+                    "}"
+                el.string = "Examples…"
+                el.insert_before(soup.new_tag("br"))
+
+            # transform div structure into 'examples'-like list
+            for el in sense.find_all(class_="entrySynList"):
+                el.name = "ul"
+                moretag = soup.new_tag("a")
+                moretag.string = "Synonyms…"
+                moretag["href"] = "#"
+                moretag["protect"] = ""  # mark
+                # important: use the 'ul' at index 1 for synonyms
+                moretag["onclick"] = \
+                    "var els = this.parentNode.getElementsByTagName('ul')[1].getElementsByTagName('li');" + \
+                    "for(var c = 0; c < els.length; c++){" + \
+                        "if (els[c].getAttribute('data-toggled') == 'false') {" + \
+                            "els[c].style = 'display: none;';" + \
+                            "els[c].setAttribute('data-toggled', 'true');" + \
+                        "} else {" + \
+                            "els[c].style = '';" + \
+                            "els[c].setAttribute('data-toggled', 'false');" + \
+                        "}" + \
+                    "}"
+
+                el.parent.a.insert_after(moretag)
+                moretag.insert_before(soup.new_tag("br"))
+
+                for syno in el.find_all("div"):
+                    syno.name = "li"
+
+                #el.insert_before(soup.new_tag("br"))
+
+            # simplify structure
+            for el in sense.find_all("span") + \
+                    sense.find_all("div") + \
+                    sense.find_all("a"):
+                # skip "more" tags
+                if "onclick" not in el.attrs:
+                    el.unwrap()
+
+            # remove unused classes
+            for el in sense.find_all():
+                del el["class"]
+            del sense["class"]
+
+            # hide examples
+            for el in sense.find_all("li"):
+                el["style"] = "display: none;"
+
+            entry += str(sense) #.prettify(formatter="html")
+            #entry += "<br/>"
+        entry += "<br/>"
+
+    # ugly hack. Why is this necessary:
+    #- for el in sense.find_all("a", class_="moreInformationExemples"):
+    #- AttributeError: 'NoneType' object has no attribute 'next_element'
+    soup = BeautifulSoup(entry, "html.parser")
+    for el in soup.find_all("li"):
+        if "Get more examples" == el.get_text() or \
+           "View synonyms" in el.get_text() or \
+                el.get_text() == "":
+            el.extract()
+
+    text = str(soup)
+    # another hack, I'm out of time
+    text = re.sub(r"(\d)([A-Za-z])", r"\1 \2", text)
+    return text
+
+#lookup("clerk")
+print(lookup("smudge"))