[Fix #529] Ignore month case in parsing

tidyverse · May 6, 2017 · c5c3ae3 · c5c3ae3
1 parent 5fe60e0
commit c5c3ae3
Show file tree

Hide file tree

Showing 4 changed files with 13 additions and 7 deletions.
diff --git a/NEWS.md b/NEWS.md
@@ -2,6 +2,7 @@ Version 1.6.0.9000
 ==================
 
 ### NEW FEATURES
+* [#529](https://github.com/hadley/lubridate/issues/529) Internal parser now ignores the case of alpha months (B format)
 * [#535](https://github.com/hadley/lubridate/issues/535) Rounding to `season` is now supported.
 * [#536](https://github.com/hadley/lubridate/issues/536) `as_date` and `as_datetime` now understand character vectors.
 * [#508](https://github.com/hadley/lubridate/pull/508) New parameter `locale` in `month`. Labels of the returned factors (when `label=TRUE`) now respect current locale.
@@ -11,7 +12,6 @@ Version 1.6.0.9000
 * [#401](https://github.com/hadley/lubridate/issues/401) New parameter `locale` in `wday`. Labels of the returned factors (when `label=TRUE`) now respect current locale.
 
 ### BUG FIXES
-
 * [#530](https://github.com/hadley/lubridate/issues/530) `parse_date_time` now throw warnings only for actual parsing errors (input with all NAs are silent)
 * [#534](https://github.com/hadley/lubridate/issues/534) Fix arithmetics with large numbers
 * [#507](https://github.com/hadley/lubridate/issues/507) Period and duration parsers now understand 0 units.

diff --git a/src/tparse.c b/src/tparse.c
@@ -34,8 +34,8 @@
 
 
 static const char ltnames[][5] = {"sec", "min", "hour", "mday", "mon", "year"};
-static const char *en_months[] = {"January", "February","March","April","May","June",
-                                  "July","August","September","October","November","December"};
+static const char *en_months[] = {"january", "february","march","april","may","june",
+                                  "july","august","september","october","november","december"};
 
 // increment **c and return month ix in 1..12 if parsing was successful, 0 if not.
 int parse_alpha_month(const char **c){

diff --git a/src/utils.c b/src/utils.c
@@ -1,6 +1,7 @@
 #include "constants.h"
 #include "utils.h"
 #include <stdio.h>
+#include <ctype.h>
 
 // return adjustment (in seconds) due to leap years
 // y: years after (positive) or before (negative) 2000-01-01 00:00:00
@@ -59,10 +60,10 @@ int parse_int (const char **c, const int N, const int strict) {
 }
 
 
-// Increment **c and return index in 0..length(strings) if match was found, -1
+// Increment *c and return index in 0..length(strings) if match was found, -1
 // if not.
 //
-// - **c: pointer to a character in a C string (automatically incremented)
+// - *c: pointer to a character in a C string (incremented by side effect)
 // - *stings: pointer to an array of C strings to be matched to.
 // - strings_len: length of strings array.
 int parse_alphanum(const char **c, const char **strings, const int strings_len){
@@ -83,7 +84,7 @@ int parse_alphanum(const char **c, const char **strings, const int strings_len){
       if (track[i]){
         // keep going while at least one valid track
         if (strings[i][j]){
-          if(**c == strings[i][j]){
+          if(**c == strings[i][j] || tolower(**c) == strings[i][j]){
             out = i;
             go = 1;
           } else {

diff --git a/tests/testthat/test-parsers.R b/tests/testthat/test-parsers.R
@@ -924,7 +924,12 @@ test_that("`parse_date_time` parses heterogeneous formats with `exact=TRUE`", {
                           tz = "UTC"))
 })
 
-
+test_that("parser ignores case", {
+  ref <- ymd_hms("2016-01-04 07:40:00", "2016-01-04 07:40:00", "2016-01-04 07:40:00 UTC")
+  dts <- c("04jan2016:07:40:00", "04JAN2016:07:40:00", "04Jan2016:07:40:00")
+  expect_equal(ref, parse_date_time2(dts, "dBYHMS"))
+  expect_equal(ref, dmy_hms(dts))
+})
 
 ## library(microbenchmark)
 ## library(lubridate)