From: Jeremy Stanley <fungi@yuggoth.org>
Date: Fri, 17 Dec 2021 16:29:38 +0000 (+0000)
Subject: Force UTF-8 locale when reading configs and data
X-Git-Tag: 2.4.3~5
X-Git-Url: https://yuggoth.org/gitweb?a=commitdiff_plain;h=455afdc07dc3b38be60b2386c8c083664fa2815a;p=weather.git

Force UTF-8 locale when reading configs and data

Apparently, Python on Windows defaults to assuming CP1252 encoding
unless otherwise specified, as opposed to the UTF-8 assumption made
on POSIX platforms. Since our configuration and data files are
expected to always use UTF-8 encoding, be clear in the
ConfigParser.read() calls about that. We only do this under Python
3.x, as that method doesn't have an encoding parameter in 2.7.

Thanks to Lance Bermudez for reporting this.
---

diff --git a/weather.py b/weather.py
index dbba754..d9f1931 100644
--- a/weather.py
+++ b/weather.py
@@ -608,7 +608,11 @@ def get_config():
         "weatherrc"
         ]
     for rcfile in rcfiles:
-        if os.access(rcfile, os.R_OK): config.read(rcfile)
+        if os.access(rcfile, os.R_OK):
+            if pyversion("3"):
+                config.read(rcfile, encoding="utf-8")
+            else:
+                config.read(rcfile)
     for section in config.sections():
         if section != section.lower():
             if config.has_section(section.lower()):
@@ -644,7 +648,10 @@ def integrate_search_cache(config, cachedir, setpath):
             pass
         return config
     scache = configparser.ConfigParser()
-    scache.read(scache_fn)
+    if pyversion("3"):
+        scache.read(scache_fn, encoding="utf-8")
+    else:
+        scache.read(scache_fn)
     for section in scache.sections():
         if not config.has_section(section):
             config.add_section(section)
@@ -742,7 +749,10 @@ def guess(
                     gzip.open(datafile).read().decode("utf-8") )
             else: stations.readfp( gzip.open(datafile) )
         else:
-            stations.read(datafile)
+            if pyversion("3"):
+                stations.read(datafile, encoding="utf-8")
+            else:
+                stations.read(datafile)
     else:
         message = "%s error: can't find \"%s\" data file\n" % (
             os.path.basename( sys.argv[0] ),
@@ -760,7 +770,10 @@ def guess(
                 zones.read_string( gzip.open(datafile).read().decode("utf-8") )
             else: zones.readfp( gzip.open(datafile) )
         else:
-            zones.read(datafile)
+            if pyversion("3"):
+                zones.read(datafile, encoding="utf-8")
+            else:
+                zones.read(datafile)
     else:
         message = "%s error: can't find \"%s\" data file\n" % (
             os.path.basename( sys.argv[0] ),
@@ -787,7 +800,10 @@ def guess(
                         gzip.open(datafile).read().decode("utf-8") )
                 else: airports.readfp( gzip.open(datafile) )
             else:
-                airports.read(datafile)
+                if pyversion("3"):
+                    airports.read(datafile, encoding="utf-8")
+                else:
+                    airports.read(datafile)
         else:
             message = "%s error: can't find \"%s\" data file\n" % (
                 os.path.basename( sys.argv[0] ),
@@ -876,7 +892,10 @@ def guess(
                         gzip.open(datafile).read().decode("utf-8") )
                 else: zctas.readfp( gzip.open(datafile) )
             else:
-                zctas.read(datafile)
+                if pyversion("3"):
+                    zctas.read(datafile, encoding="utf-8")
+                else:
+                    zctas.read(datafile)
         else:
             message = "%s error: can't find \"%s\" data file\n" % (
                 os.path.basename( sys.argv[0] ),
@@ -934,7 +953,10 @@ def guess(
                         gzip.open(datafile).read().decode("utf-8") )
                 else: places.readfp( gzip.open(datafile) )
             else:
-                places.read(datafile)
+                if pyversion("3"):
+                    places.read(datafile, encoding="utf-8")
+                else:
+                    places.read(datafile)
         else:
             message = "%s error: can't find \"%s\" data file\n" % (
                 os.path.basename( sys.argv[0] ),
@@ -1151,7 +1173,10 @@ def guess(
             )
         try:
             scache_existing = configparser.ConfigParser()
-            scache_existing.read(scache_fn)
+            if pyversion("3"):
+                scache_existing.read(scache_fn, encoding="utf-8")
+            else:
+                scache_existing.read(scache_fn)
             if not scache_existing.has_section(search[0]):
                 scache_fd = codecs.open(scache_fn, "a", "utf-8")
                 scache_fd.writelines(search_cache)
@@ -2003,15 +2028,30 @@ def correlate():
     sys.stdout.write(message)
     sys.stdout.flush()
     airports = configparser.ConfigParser()
-    airports.read(airports_fn)
+    if pyversion("3"):
+        airports.read(airports_fn, encoding="utf-8")
+    else:
+        airports.read(airports_fn)
     places = configparser.ConfigParser()
-    places.read(places_fn)
+    if pyversion("3"):
+        places.read(places_fn, encoding="utf-8")
+    else:
+        places.read(places_fn)
     stations = configparser.ConfigParser()
-    stations.read(stations_fn)
+    if pyversion("3"):
+        stations.read(stations_fn, encoding="utf-8")
+    else:
+        stations.read(stations_fn)
     zctas = configparser.ConfigParser()
-    zctas.read(zctas_fn)
+    if pyversion("3"):
+        zctas.read(zctas_fn, encoding="utf-8")
+    else:
+        zctas.read(zctas_fn)
     zones = configparser.ConfigParser()
-    zones.read(zones_fn)
+    if pyversion("3"):
+        zones.read(zones_fn, encoding="utf-8")
+    else:
+        zones.read(zones_fn)
     qalog = []
     places_nocentroid = 0
     places_nodescription = 0