https://bugs.documentfoundation.org/show_bug.cgi?id=96572

            Bug ID: 96572
           Summary: garbage for null in non-nullable integer field
           Product: LibreOffice
           Version: 5.2.0.0.alpha0+ Master
          Hardware: x86-64 (AMD64)
                OS: All
            Status: UNCONFIRMED
          Severity: normal
          Priority: medium
         Component: Base
          Assignee: [email protected]
          Reporter: [email protected]

STR
---

( 0) In the following steps, I have inserted blanks for readability in
     the strings to be typed.  Do not actually type the blanks,

( 1) Enable experimental features.

( 2) Download and open attached fb.odb.  The enclosed Table1 has the
     first column "number" as primary key.

( 3) In pane Database, click icon <Tables>.

( 4) In pane Tables, right click Table1, and from pop-up menu select
     Open.

( 5) In window "Table 1 - fb", observe two columns and three rows of
     data, the values of "number" being (1,2,3), followed by a fourth
     empty row.

( 6) Type "<Down><Down><Down> 3 <Tab> drei <Tab>".  Program displays error
message (newlines inserted):

         Error inserting the new record

         firebird_sdbc error:
         *violation of PRIMARY or UNIQUE KEY constraint "INTEG_2" on
             table "Table1"
         *Problematic key value is ("number = 3)
         caused by
         'isc_dsql_execute'

     This message is appropriate.

( 7) Click <OK>.  Program returns focus to "Table 1 - fb"; "drei" is
     reverse-image.

( 8) Type "<Backspace> <shift>+<Tab> <Backspace>".  The fourth row of
     the table is blank.

( 9) Type "<Alt>+F S".

(10) Type "<Alt>+F C".  Program returns focus to window "fb.odb".

(11) In pane Tables, right click Table1, and from pop-up menu select
     Open.

(12) In window "Table1 - fb", observe ...

     Expected : Three rows of data followed by a fourth blank row.

     Actual : The fourth row in column "number" shows a large number
              like -100920357 or 12632430 or -724851064; there is a
              fifth blank row.


I have observed this with LibreOffice daily dbgutil bibisect version
2015-12-04 ...

    Version: 5.2.0.0.alpha0+
    Build ID: 1fd41f43eb73c373cb94d32d82c5fb7a7e243367
    Threads 1; Ver: Linux 3.2; Render: default; 

    Locale: en-CA (en_CA.UTF-8)

... running in an environment chroot'd to debian-sid, and on some
Windows daily build, so I am setting O/S=All An embedded HSQLDB
database is similarly afflicted.

Within the daily dbgutil bibisect repository started 2015-05-20, I get
from `git bisect good` ...

    f39c4096f247626f7447321154de73b0c0dbc6b7 is the first bad commit
    commit f39c4096f247626f7447321154de73b0c0dbc6b7
    Author: Miklos Vajna <[email protected]>
    Date:   Sun May 24 05:47:28 2015 +0200

        2015-05-24: source-hash-891304bb0ad3af9d8d73f947f25477abf57485a4

    :100644 100644 52f908d74b08372c4af8ab39a1d305b4340a9814
a8025630e6a3b02c8689947005f15c432a3d9124 M    build-info.txt
    :040000 040000 690e3195db7163c207fe50b4510f02058f5531f7
039a1db6db12117a8679909399f865b4dc0d0d16 M    opt

... and from `git bisect log` ...

    # bad: [c5625a3e1014766a488ca42e986be5dd3efbf051] 2015-11-24:
source-hash-e5b3ec9683e4c5761633a1a3c4819a039e796b01
    # good: [2b392af9c8f54629e3a3a98a8c92fa5af1c6722f] 2015-05-20:
source-hash-90e2dabb8d0bb5382234be776c2ad0e2d5d9e224
    git bisect start 'c5625a3' '2b392af'
    # bad: [a8d19eafb9e195a85359dd8cdc5c46c38295caa0] 2015-08-22:
source-hash-79fb61efb847405fa47235002b52ee8efad5e339
    git bisect bad a8d19eafb9e195a85359dd8cdc5c46c38295caa0
    # bad: [c9abefa2facd9d5f2989bbc2138e55970e60c4cd] 2015-07-06:
source-hash-64385c7555c6bbb089f3ab8b2a5b623cadd6ee7d
    git bisect bad c9abefa2facd9d5f2989bbc2138e55970e60c4cd
    # bad: [d3328123e5e86a74111ff9f1fcdd742973bd3c80] 2015-06-12:
source-hash-bb9dad2ef23829b2500c9f99154bca6a8ba7d49a
    git bisect bad d3328123e5e86a74111ff9f1fcdd742973bd3c80
    # bad: [f66b13d237da5d0a5ec7b5856a2c3b8224c30aa6] 2015-05-31:
source-hash-4618d15cb73d0f0c15e0f4d2a6ce9cdbe6eeef24
    git bisect bad f66b13d237da5d0a5ec7b5856a2c3b8224c30aa6
    # bad: [a63d7336fa8ea346b4deb2d4076cbe8a1cd75d65] 2015-05-25:
source-hash-b9630867d17c01ec41f6461b1e96288f3932248c
    git bisect bad a63d7336fa8ea346b4deb2d4076cbe8a1cd75d65
    # good: [33e7c549fd7aa03e831a939201aa06dd80edb2b0] 2015-05-22:
source-hash-edcd1d5a9c88455fd1d52ab25815fc8d439f31ec
    git bisect good 33e7c549fd7aa03e831a939201aa06dd80edb2b0
    # bad: [f39c4096f247626f7447321154de73b0c0dbc6b7] 2015-05-24:
source-hash-891304bb0ad3af9d8d73f947f25477abf57485a4
    git bisect bad f39c4096f247626f7447321154de73b0c0dbc6b7
    # good: [6178f2a1052f501e7b34ad168708c6744a1838eb] 2015-05-23:
source-hash-9db629b8a1fa9b63bc320f8d47594ec82511a9c5
    git bisect good 6178f2a1052f501e7b34ad168708c6744a1838eb
    # first bad commit: [f39c4096f247626f7447321154de73b0c0dbc6b7] 2015-05-24:
source-hash-891304bb0ad3af9d8d73f947f25477abf57485a4

... so I am setting keywords regression bibisected.

-- 
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
Libreoffice-bugs mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/libreoffice-bugs

Reply via email to