gtfs_df 0.11.1 → 0.12.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 70a7d05d5600baac0b36b21d0ebdd8fd0719cfe572844e44ae710c25fbf353f8
-  data.tar.gz: df2b5f47cd19bfe213644dcda42b22778acf98d8f104bf71c4e231eaa4860bba
+  metadata.gz: 8c1060a2f120fc6f620adc523d8038a05ec6896826fbb950973c14b8a406fc37
+  data.tar.gz: 6f75be7bbec96e5f75994fa130f9d599061d796ef702cad6c833d8edb873ef6b
 SHA512:
-  metadata.gz: ee94819792b5c1389783a7844ce23af40de13264c5c069e608ae9c2383254c3fdcdc09915f5c33348d204675e5e373f8c26b039ff97144be37721aa7f4bc1b49
-  data.tar.gz: 28f28bb69a62738c817dafbc36c756a96f56f9f96d4d587aab2b9dffa8df5b6948b25d6b024e70856639cc2a97c75c49ff510bb7695c825f54405bf23701582d
+  metadata.gz: c8000394c34a67284d16df01e9e236b39d074bd300d1416be8557a7db2387357e860b7c741ff3eb483293895ca16bb3936efd857489124d309ced30a06d248aa
+  data.tar.gz: 15a5d19bb1d06ab1c9f21ae0e403ce6187dd370dcf868d20b9ee1de415a2a1ecabb1ae158b5dc10bc6236f329863e661581a2f8b39d0fb9eed50bc9dbad350db

data/CHANGELOG.md

@@ -1,3 +1,13 @@
+## [0.12.0] - 2026-06-12
+
+### 🚀 Features
+
+- Allow loading a subset of tables
+
+### 🐛 Bug Fixes
+
+- Truncate ragged lines
+- Consider BOM when looking for empty files
 ## [0.11.1] - 2026-04-28
 
 ### 🐛 Bug Fixes
@@ -8,6 +18,10 @@
 
 - Validate filtering from feed with empty calendar
 - Validate unreferenced calendar dates
+
+### ⚙️ Miscellaneous Tasks
+
+- Bump version to 0.11.1
 ## [0.11.0] - 2026-03-30
 
 ### ⚙️ Miscellaneous Tasks

data/lib/gtfs_df/base_gtfs_table.rb

@@ -12,7 +12,13 @@ module GtfsDf
         elsif input.is_a?(String)
           # TODO: use `infer_schema: false` instead of `infer_schema_length` after polars release:
           # https://github.com/ankane/ruby-polars/blob/master/CHANGELOG.md#100-unreleased
-          df = Polars.read_csv(input, infer_schema_length: 0, encoding: "utf8-lossy")
+          df = Polars
+            .read_csv(
+              input,
+              infer_schema_length: 0,
+              encoding: "utf8-lossy",
+              truncate_ragged_lines: true
+            )
             .rename(->(col) { col.strip })
 
           # Strip out empty lines. Unfortunately read_csv does not support the drop_empty_rows

data/lib/gtfs_df/reader.rb

@@ -6,13 +6,13 @@ module GtfsDf
     #
     # @param zip_path [String] Path to the GTFS zip file
     # @param parse_times [Boolean] Whether to parse time fields to seconds since midnight (default: false)
+    # @param relevant_files [Array<String>] A list of file names, useful to avoid loading tables you don't care about.
     # @return [Feed] The loaded GTFS feed
-    def self.load_from_zip(zip_path, parse_times: false)
+    def self.load_from_zip(zip_path, parse_times: false, relevant_files: nil)
       data = nil
 
-      relevant_files = GtfsDf::Feed::GTFS_FILES
-        .map { |name| "#{name}.txt" }
-        .to_set
+      relevant_files ||= GtfsDf::Feed::GTFS_FILES.map { |name| "#{name}.txt" }
+      relevant_files = relevant_files.to_set
 
       seen = {}
 
@@ -37,7 +37,7 @@ module GtfsDf
           end
         end
 
-        data = load_from_dir(tmpdir, parse_times: parse_times)
+        data = load_from_dir(tmpdir, parse_times:, relevant_files:)
       end
 
       data
@@ -47,12 +47,17 @@ module GtfsDf
     #
     # @param dir_path [String] Path to the GTFS directory
     # @param parse_times [Boolean] Whether to parse time fields to seconds since midnight (default: false)
+    # @param relevant_files [Array<String>] A list of file names, useful to avoid loading tables you don't care about.
     # @return [Feed] The loaded GTFS feed
-    def self.load_from_dir(dir_path, parse_times: false)
+    def self.load_from_dir(dir_path, parse_times: false, relevant_files: nil)
+      relevant_files ||= GtfsDf::Feed::GTFS_FILES.map { |name| "#{name}.txt" }
+      relevant_files = relevant_files.to_set
+
       data = {}
       GtfsDf::Feed::GTFS_FILES.each do |gtfs_file|
-        path = File.join(dir_path, "#{gtfs_file}.txt")
-        next unless File.exist?(path)
+        basename = "#{gtfs_file}.txt"
+        path = File.join(dir_path, basename)
+        next unless relevant_files.include?(basename) && File.exist?(path)
 
         data[gtfs_file] = data_frame(gtfs_file, path)
       end
@@ -66,7 +71,11 @@ module GtfsDf
     end
 
     private_class_method def self.has_header?(zip_entry)
-      zip_entry.get_input_stream.readline.strip != ""
+      zip_entry
+        .get_input_stream
+        .readline
+        .delete_prefix("\xEF\xBB\xBF".b) # BOM
+        .strip != ""
     rescue
       false
     end

data/lib/gtfs_df/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module GtfsDf
-  VERSION = "0.11.1"
+  VERSION = "0.12.0"
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: gtfs_df
 version: !ruby/object:Gem::Version
-  version: 0.11.1
+  version: 0.12.0
 platform: ruby
 authors:
 - David Mejorado