1. 10 Apr, 2019 1 commit
  2. 05 Apr, 2019 1 commit
  3. 07 Jan, 2019 1 commit
  4. 13 Dec, 2018 1 commit
  5. 03 Dec, 2018 1 commit
  6. 10 Oct, 2018 1 commit
  7. 06 Sep, 2018 1 commit
  8. 06 Jul, 2018 1 commit
  9. 27 Jun, 2018 1 commit
  10. 28 May, 2018 1 commit
  11. 27 May, 2018 1 commit
  12. 19 Apr, 2018 1 commit
    • totten's avatar
      (NFC) Update version in header · fee14197
      totten authored
      This is a simple administrative update to the headers. It was generated with the command:
      
      ```
      rgrep '| CiviCRM version 4.7' CRM/ Civi ang api bin extern install/ settings/ templates -l \
        | xargs sed -i'' "s/| CiviCRM version 4.7/| CiviCRM version 5  /g"
      ```
      
      Tthe inclusion of `|` aimed to avoid matching any non-header text (e.g. inline docs that
      mentioned the version incidentally). But then I did a looser search and for just
      
      ```
      rgrep 'CiviCRM version 4.7'
      ````
      
      and manually patched the remainder.
      
      Note: I'm not really keen on doing this every month, so I relaxed the header
      statement -- instead of `CiviCRM version 5.0`, it's just `CiviCRM version 5`.
      fee14197
  13. 23 Feb, 2018 1 commit
  14. 25 Jan, 2018 1 commit
    • eileen's avatar
      CRM-21707 use metadata to validate 'parents' field on group.parents · 1678a63b
      eileen authored
      There was a test attempting to pass
      
      [1 => 'Test Group'] which was getting through to the BAO but I feel should
      have been blocked at the api level as Test Group did not exist. We have existing code to validate against pseudoconstants but this was not kicking in
      because
      a) there was no pseudoconstant defined for the field
      b) we were giving TEXT fields less validation than STRING fields
      (the assumption looks like it was that they were html fields but that is
      not consistent & IMHO non of the additional checks are harmful.)
      c) the getoptions was failing because it was passing 'validate' in
      as a param to the pseudoconstant once set up, I handled that in
      the pseudoconstant function.
      
      I also altered the test to see [1 => 'non existent group'] as a fail
      (rather than converting it to group = 1 & succeed.
      
      There are a couple of follow on thoughts I have
      1) I think we should xss check keys as well as values - will add that as a commit
      2) there may be a very very hypothetical hole in _civicrm_api3_api_match_pseudoconstant_value - if  is empty it returns early. I think that used
      to be valid as some psuedoconstants were not resolveable but now
      I think if options is empty it means there are no valid options &
      it should be rejected. Note that these strings are still xss checked.
      1678a63b
  15. 16 Jan, 2018 1 commit
  16. 04 Dec, 2017 1 commit
  17. 22 Nov, 2017 1 commit
  18. 21 Nov, 2017 1 commit
  19. 03 May, 2017 1 commit
  20. 19 Apr, 2017 1 commit
  21. 10 Jan, 2017 1 commit
  22. 02 Jan, 2017 1 commit
  23. 14 Dec, 2016 1 commit
  24. 13 Nov, 2016 1 commit
  25. 12 Oct, 2016 1 commit
  26. 13 Aug, 2016 3 commits
  27. 08 Aug, 2016 1 commit
    • eileenmcnaugton's avatar
      CRM-19126 load previous values selectively for tax · 2912ed09
      eileenmcnaugton authored
      This includes moving the setting of tax values from the api to the BAO.
      
      Unfortunately an awful lot of the tax work was done in the form layer & it would
      be a big job to figure out which parts can be left to the BAO
      2912ed09
  28. 07 Aug, 2016 1 commit
  29. 15 Jun, 2016 1 commit
  30. 08 Apr, 2016 1 commit
  31. 21 Mar, 2016 1 commit
  32. 17 Mar, 2016 1 commit
  33. 03 Mar, 2016 1 commit
  34. 21 Jan, 2016 1 commit
  35. 28 Dec, 2015 1 commit
  36. 21 Dec, 2015 3 commits