From 97e060fa0cf21d3be385b386f355702dc0a765ae Mon Sep 17 00:00:00 2001
From: tom5079 <tom5079@naver.com>
Date: Mon, 5 Oct 2020 18:06:49 +0900
Subject: [PATCH] (WIP) Added update script

---
 .gitignore               |  1 +
 en/tags.json             |  3 ---
 ko.json                  |  0
 scripts/requirements.txt |  2 ++
 scripts/tags.py          | 48 ++++++++++++++++++++++++++++++++++++++++
 scripts/update.py        |  0
 scripts/update.sh        |  0
 template.json            |  0
 8 files changed, 51 insertions(+), 3 deletions(-)
 create mode 100644 .gitignore
 delete mode 100644 en/tags.json
 create mode 100644 ko.json
 create mode 100644 scripts/requirements.txt
 create mode 100644 scripts/tags.py
 create mode 100644 scripts/update.py
 create mode 100644 scripts/update.sh
 create mode 100644 template.json

diff --git a/.gitignore b/.gitignore
new file mode 100644
index 00000000..dbe9c82b
--- /dev/null
+++ b/.gitignore
@@ -0,0 +1 @@
+.vscode/
\ No newline at end of file
diff --git a/en/tags.json b/en/tags.json
deleted file mode 100644
index c5e02bfb..00000000
--- a/en/tags.json
+++ /dev/null
@@ -1,3 +0,0 @@
-{
-    "test": "test"
-}
\ No newline at end of file
diff --git a/ko.json b/ko.json
new file mode 100644
index 00000000..e69de29b
diff --git a/scripts/requirements.txt b/scripts/requirements.txt
new file mode 100644
index 00000000..a98ae430
--- /dev/null
+++ b/scripts/requirements.txt
@@ -0,0 +1,2 @@
+requests
+beautifulsoup4
\ No newline at end of file
diff --git a/scripts/tags.py b/scripts/tags.py
new file mode 100644
index 00000000..5d9d626e
--- /dev/null
+++ b/scripts/tags.py
@@ -0,0 +1,48 @@
+# Returns a list of all tags sorted by the number of tags
+
+import re
+import json
+
+import requests
+from bs4 import BeautifulSoup
+
+indices = ['123'] + [chr(ord('a')+i) for i in range(26)]
+tags = dict()
+
+count_regex = re.compile(r".+\((\d+)\)$")
+
+for index in indices:
+    url = f'https://hitomi.la/alltags-{index}.html'
+
+    soup = BeautifulSoup(requests.get(url).text, 'html.parser')
+
+    for item in soup.select('.content li'):
+        tag = item.a.text
+        count = int(count_regex.match(item.text).group(1))
+
+        tags[tag] = count
+    
+    break
+
+tag_regex = re.compile(r".+:(.+)$")
+def clean(tag):
+    match = tag_regex.match(tag)
+
+    if match:
+        return match.group(1)
+    else:
+        return tag
+
+temp = dict()
+for k, v in tags.items():
+    tag = clean(k)
+
+    if tag in temp:
+        if v > temp[tag]:
+            temp[tag] = v
+    else:
+        temp[tag] = v
+
+tags = sorted(temp, key=temp.get, reverse=True)
+
+print(json.dumps(tags, indent=4))
\ No newline at end of file
diff --git a/scripts/update.py b/scripts/update.py
new file mode 100644
index 00000000..e69de29b
diff --git a/scripts/update.sh b/scripts/update.sh
new file mode 100644
index 00000000..e69de29b
diff --git a/template.json b/template.json
new file mode 100644
index 00000000..e69de29b