Commit graph

10 commits

Author SHA1 Message Date
Mickey Rose
c21778fdfc merge mapnik-index::process_csv_file and csv_datasource::parse_csv
- the function in plugin was already configurable via flags, and only
  contained two un-conditioned blocks that process_csv_file didn't have

- so I extracted the common parts into a separate function (in a class
  holding the flags and state), process_csv_file calls it with default
  flags, plugin sets them from params

- removed namespace ::detail, moving stuff that was used outside
  csv_utils to ::csv_utils, and the rest to ::csv_utils::detail
2016-02-26 20:28:42 +01:00
Dane Springmeyer
8813e73cfc rename SHAPE_MEMORY_MAPPED_FILE to MAPNIK_MEMORY_MAPPED_FILE 2015-10-16 13:34:53 -07:00
artemp
178e39e19a make separator single character and simplify/optimise csv_grammar 2015-10-05 15:23:20 +01:00
artemp
1d320b7133 make quote single char 2015-10-02 12:20:54 +01:00
artemp
0de6d36000 use memory mapped file by default for both index and data parsing (improves loading times by ~50%) - default on non-windows platforms 2015-09-09 11:56:21 +02:00
artemp
8709fb6f7c CSV - optimise parsing by providing num_columns hint 2015-08-24 16:35:32 +02:00
artemp
6c9257a915 add parse_line accepting iterator range and avoid string ctor 2015-08-24 15:41:04 +02:00
artemp
c9d1d51b8a simplify + factor out properties parsing logic 2015-08-24 14:13:13 +02:00
artemp
2477d8764e keep on untangling spaghetti
* implement standalone ignore case equality to avoid copying
* fix various logic shortcommings
2015-08-24 12:23:59 +02:00
artemp
4babec802a CSV - implement spatial index access to features on disk + preserve support for inline data (work-in-progress) 2015-08-21 13:52:42 +02:00