From e1acf5b6e8cbbcf281d344f3d16badd535cd6d37 Mon Sep 17 00:00:00 2001
From: lonkaars <loek@pipeframe.xyz>
Date: Sat, 8 Apr 2023 13:33:25 +0200
Subject: implement cache in import script

---
 dbinit/.gitignore |  1 +
 dbinit/import.py  | 54 +++++++++++++++++++++++++++++++++++++++++++++---------
 2 files changed, 46 insertions(+), 9 deletions(-)

diff --git a/dbinit/.gitignore b/dbinit/.gitignore
index e95719f..53afa4c 100644
--- a/dbinit/.gitignore
+++ b/dbinit/.gitignore
@@ -1,3 +1,4 @@
 full.sql
 base.sql
 data.sql
+.cache.*
diff --git a/dbinit/import.py b/dbinit/import.py
index bf81bfb..f79be98 100755
--- a/dbinit/import.py
+++ b/dbinit/import.py
@@ -2,23 +2,59 @@
 
 import os
 import sys
+import json
 import urllib.request
+import urllib.parse
+import re
+from pathlib import Path
 
-global YEAR
 USE_LOCAL_CACHE = True
-API_BASE_URL = "https://ergast.com/"
+API_BASE_URL = "https://ergast.com/api/"
 
-def api_request(endpoint):
-  conn = urllib.request.urlopen(API_BASE_URL + endpoint)
-  return conn.read()
+def url2localpath(input_str):
+  input_str = re.sub(r"(https)|(ergast\.com)|(api)|[:/?&=]", ".", input_str)
+  input_str = re.sub(r"\.+", ".", ".cache." + input_str)
+  return input_str
 
-def main():
-  print(f"fetching year {YEAR}")
+def cache_check(url):
+  if not USE_LOCAL_CACHE: return None
+  path = url2localpath(url)
+  if not Path(path).is_file(): return None
+  res = ""
+  with open(path, "r") as f:
+    res = f.read()
+    f.close()
+  return res
+
+def cache_save(url, response):
+  if not USE_LOCAL_CACHE: return
+  path = url2localpath(url)
+  with open(path, "w+") as f:
+    f.write(response)
+    f.close()
+
+def api_request(endpoint, options={}):
+  url = urllib.parse.urljoin(API_BASE_URL, endpoint + ".json")
+  options_str = urllib.parse.urlencode(options)
+  if options_str: url += "?" + options_str
+
+  cache_hit = cache_check(url)
+  if cache_hit != None:
+    return cache_hit
+  
+  print("actually requesting api")
+  conn = urllib.request.urlopen(url)
+  res = conn.read().decode("utf-8")
+  cache_save(url, res)
+  return json.loads(res)
+
+def main(year):
+  print(f"fetching year {year}")
+  print(api_request("f1/2018/1"))
 
 if __name__ == "__main__":
   if len(sys.argv) < 2:
     print("please provide a year to fetch f1 data from")
     exit(1)
-  YEAR = int(sys.argv[1])
-  main()
+  main(sys.argv[1])
 
-- 
cgit v1.2.3