Rename to kanji_info_fetcher

4 files changed, 226 insertions, 0 deletions

diff --git a/kanji_info_fetcher/__init__.py b/kanji_info_fetcher/__init__.py
new file mode 100644
index 0000000..c4b95bb
--- /dev/null
+++ b/kanji_info_fetcher/__init__.py
@@ -0,0 +1,42 @@
+import argparse
+import kanji_info_fetcher.io
+import kanji_info_fetcher.jisho_client
+import kanji_info_fetcher.wadoku_client
+
+__version__ = '0.1.0'
+
+DEFAULT_FORMAT = 'kanji,english'
+
+def main():
+  parser = argparse.ArgumentParser(description='Convert course vocab into Anki deck')
+  parser.add_argument('--format', metavar='format', type=str, default=DEFAULT_FORMAT,
+                      help='comma-separated row format, e.g. kanji,hiragana,english,readings')
+  parser.add_argument('file', metavar='file', type=str,
+                      help='vocabulary file path')
+
+  args = parser.parse_args()
+  file_path = args.file
+  row_format = args.format
+
+  print("row_format:", row_format)
+
+  # read entries from file
+  print(f"[STATUS] Reading file {file_path}")
+  entries = io.read_from_file(file_path, row_format)
+
+  # fetch translations from Jisho
+  print(f"[STATUS] Fetching English translations from Jisho")
+  entries = jisho_client.inject_english_translation(entries)
+
+  # fetch kanji meanings from Jisho
+  print(f"[STATUS] Fetching kanji meaning data from Jisho")
+  entries = jisho_client.inject_kanji_meanings(entries)
+
+  # fetch hiragana readings from Wadoku w/ pitch accents
+  print(f"[STATUS] Fetching reading data from Wadoku")
+  entries = wadoku_client.inject_hiragana_with_accents(entries)
+
+  # write entries to file
+  out_file_name = file_path.replace(".csv", ".out.csv") if ".csv" in file_path else f"{file_path}.out"
+  print(f"[STATUS] Writing result to file {out_file_name}")
+  io.write_to_file(out_file_name, entries)
diff --git a/kanji_info_fetcher/io.py b/kanji_info_fetcher/io.py
new file mode 100644
index 0000000..90c9507
--- /dev/null
+++ b/kanji_info_fetcher/io.py
@@ -0,0 +1,50 @@
+from os import path
+from dataclasses import dataclass
+
+@dataclass
+class Entry:
+  kanji: str
+  hiragana: str
+  english: str
+  meanings: str
+
+def read_from_file(path_str: str, row_format: str) -> list:
+  p = path.join(path_str)
+  lines = []
+
+  format_segments = row_format.split(",")
+
+  with open(p, "r") as f:
+    lines = f.readlines()
+
+  entries = []
+  for line in lines:
+    if len(line.strip()) == 0:
+      continue
+
+    value_dict = {}
+
+    row_values = line.split(",", len(format_segments) - 1)
+    for i, format_segment in enumerate(format_segments):
+      value_dict[format_segment] = row_values[i]
+
+    entries.append(Entry(
+      kanji=value_dict.get("kanji", "").strip(),
+      hiragana=value_dict.get("hiragana", "").strip(),
+      english=value_dict.get("english", "").strip(),
+      meanings=value_dict.get("meanings", "").strip()
+    ))
+
+  return entries
+
+def _entry_to_row(entry: Entry, delimiter: str = "§") -> str:
+  s = delimiter.join([entry.kanji, entry.hiragana, entry.english, entry.meanings])
+  return f"{s}\n"
+
+def write_to_file(path_str: str, entries: list):
+  p = path.join(path_str)
+
+  lines = map(_entry_to_row, entries)
+
+  with open(p, "w") as f:
+    f.writelines(lines)
diff --git a/kanji_info_fetcher/jisho_client.py b/kanji_info_fetcher/jisho_client.py
new file mode 100644
index 0000000..e74412e
--- /dev/null
+++ b/kanji_info_fetcher/jisho_client.py
@@ -0,0 +1,92 @@
+from bs4 import BeautifulSoup
+import requests
+import urllib.parse
+
+JISHO_BASE_URL = "https://jisho.org/search"
+
+def _meaning_tag_to_abbrev(tag: str) -> str:
+  if "Godan" in tag: return "G1"
+  if "Ichidan" in tag: return "G2"
+  if "Suru" in tag: return "G3"
+  if "Noun" in tag: return "n."
+  if "I-adjective" in tag: return "い-adj."
+  if "Na-adjective" in tag: return "な-adj."
+  if "No-adjective" in tag: return "の-adj."
+  if "Adverb" in tag: return "adv."
+  return tag
+
+def inject_kanji_meanings(entries: list) -> list:
+  entries_with_meanings = []
+
+  for index, entry in enumerate(entries):
+    kanji = entry.kanji
+
+    if len(entry.meanings) > 0:
+      print(f"Meaning data exists for #{index}: {kanji}. Skipping...")
+      entries_with_meanings.append(entry)
+      continue
+
+    print(f"Fetching meaning data from Jisho for #{index}: {kanji}")
+
+    resource_part = urllib.parse.quote_plus(f"{kanji}#kanji")
+    url = f"{JISHO_BASE_URL}/{resource_part}"
+    resp = requests.get(url)
+    resp.raise_for_status()
+
+    soup = BeautifulSoup(resp.text, "html.parser")
+    container = soup.select_one("div#page_container")
+
+    meanings = container.select(".kanji-details__main-meanings")
+    kanji_meaning_texts = map(lambda e: e.text.strip(), meanings)
+
+    kanji_characters = container.select(".character")
+    kanji_character_texts = map(lambda e: e.text.strip(), kanji_characters)
+
+    result_tuples = zip(kanji_character_texts, kanji_meaning_texts)
+    result_strs = map(lambda t: " ".join(t), result_tuples)
+    result = "<br>".join(result_strs)
+
+    entry.meanings = result
+    entries_with_meanings.append(entry)
+
+  return entries_with_meanings
+
+def inject_english_translation(entries: list) -> list:
+  entries_with_translations = []
+
+  for index, entry in enumerate(entries):
+    kanji = entry.kanji
+
+    if len(entry.english) > 0:
+      print(f"English translation exists for #{index}: {kanji}. Skipping...")
+      entries_with_translations.append(entry)
+      continue
+
+    print(f"Fetching translation from Jisho for #{index}: {kanji}")
+
+    resource_part = urllib.parse.quote_plus(f"{kanji}")
+    url = f"{JISHO_BASE_URL}/{resource_part}"
+    resp = requests.get(url)
+    resp.raise_for_status()
+
+    soup = BeautifulSoup(resp.text, "html.parser")
+    container = soup.select_one("div#page_container")
+
+    meaning_wrapper = container.select(".meanings-wrapper")[0]
+
+    meaning_tags = meaning_wrapper.select(".meaning-tags")
+    meaning_tags_texts = map(lambda e: e.text.strip(), meaning_tags)
+    meaning_tag_abbrevs = map(_meaning_tag_to_abbrev, meaning_tags_texts)
+
+    meaning_translations = meaning_wrapper.select(".meaning-wrapper .meaning-meaning")
+    meaning_translations_texts = map(lambda e: e.text.strip(), meaning_translations)
+
+    meanings_zipped = zip(meaning_tag_abbrevs, meaning_translations_texts)
+
+    first_translation_pair = list(meanings_zipped)[0]
+    result = f"{first_translation_pair[0]} {first_translation_pair[1]}"
+
+    entry.english = result
+    entries_with_translations.append(entry)
+
+  return entries_with_translations
diff --git a/kanji_info_fetcher/wadoku_client.py b/kanji_info_fetcher/wadoku_client.py
new file mode 100644
index 0000000..550f801
--- /dev/null
+++ b/kanji_info_fetcher/wadoku_client.py
@@ -0,0 +1,42 @@
+from bs4 import BeautifulSoup
+import requests
+import urllib.parse
+
+WADOKU_BASE_URL = "https://www.wadoku.de/search"
+
+def inject_hiragana_with_accents(entries: list) -> list:
+  entries_with_readings = []
+
+  for index, entry in enumerate(entries):
+    kanji = entry.kanji.replace("~", "")
+
+    if len(entry.hiragana) > 0:
+      print(f"Hiragana exists for #{index}: {kanji}. Skipping...")
+      entries_with_readings.append(entry)
+      continue
+
+    # Remove nasty characters from kanji since Wadoku
+    # fails to find anything otherwise
+    print(f"Fetching hiragana data for #{index}: {kanji}")
+
+    resource_part = urllib.parse.quote_plus(kanji)
+    url = f"{WADOKU_BASE_URL}/{resource_part}"
+    resp = requests.get(url)
+    resp.raise_for_status()
+
+    soup = BeautifulSoup(resp.text, "html.parser")
+    container = soup.select_one("section#content")
+
+    readings = container.select("span.reading")
+    first_reading = str(readings[0])
+
+    # Remove extra characters added by Wadoku
+    # https://www.wadoku.de/wiki/display/WAD/Hinweise+zur+Notation
+    result = first_reading \
+      .replace("~", "") \
+      .replace("|", "") \
+      .replace("・", "")
+    entry.hiragana = result
+    entries_with_readings.append(entry)
+
+  return entries_with_readings