Mercurial > hg > xemacs-beta

comparison src/search.c @ 4897:91a023144e72

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
fix longstanding search bug involving searching for Control-1 chars -------------------- ChangeLog entries follow: -------------------- src/ChangeLog addition: 2010-01-29 Ben Wing <ben@xemacs.org> * search.c (boyer_moore): Fix longstanding bug involving searching for Control-1 chars; code was trying to directly extract the last byte in the textual representation of a char from an Ichar (and doing it in a buggy fashion) rather than just converting the Ichar to text and looking at the last byte. tests/ChangeLog addition: 2010-01-29 Ben Wing <ben@xemacs.org> * automated/search-tests.el: New file. * automated/search-tests.el: * automated/case-tests.el: * automated/case-tests.el (pristine-case-table): Removed. * automated/case-tests.el (uni-mappings): * automated/lisp-tests.el: * automated/regexp-tests.el: Extract some search-related code from case-tests and regexp-tests and move to search-tests. Move some regexp-related code from lisp-tests to regexp-tests. Write a comment trying to express the proper division of labor between case-tests, search-tests and regexp-tests. Add a new test for the Control-1 search bug. Fix a buggy test in the Unicode torture-test section of case-tests.el.
author Ben Wing <ben@xemacs.org>
date Fri, 29 Jan 2010 20:57:42 -0600
parents 69b803c646cd
children 7504864a986c
comparison
equal deleted inserted replaced
4896:a7ab1d6ff301 4897:91a023144e72
1777 if (i == dirlen) 1777 if (i == dirlen)
1778 i = infinity; 1778 i = infinity;
1779 if (!NILP (trt)) 1779 if (!NILP (trt))
1780 { 1780 {
1781 #ifdef MULE 1781 #ifdef MULE
1782 Ichar ch, untranslated; 1782 Ichar ch = -1, untranslated;
1783 Ibyte byte;
1783 int this_translated = 1; 1784 int this_translated = 1;
1784 1785
1785 /* Is *PTR the last byte of a character? */ 1786 /* Is *PTR the last byte of a character? */
1786 if (pat_end - ptr == 1 || ibyte_first_byte_p (ptr[1])) 1787 if (pat_end - ptr == 1 || ibyte_first_byte_p (ptr[1]))
1787 { 1788 {
1827 know from the canon table that untranslated maps to 1828 know from the canon table that untranslated maps to
1828 starting_ch and that untranslated has the correct value 1829 starting_ch and that untranslated has the correct value
1829 for charset_base.) */ 1830 for charset_base.) */
1830 assert (1 == count || starting_ch != ch); 1831 assert (1 == count || starting_ch != ch);
1831 } 1832 }
1833 {
1834 Ibyte tmp[MAX_ICHAR_LEN];
1835 Bytecount chlen;
1836
1837 chlen = set_itext_ichar (tmp, ch);
1838 byte = tmp[chlen - 1];
1839 }
1832 } 1840 }
1833 else 1841 else
1834 { 1842 {
1835 ch = *ptr; 1843 byte = *ptr;
1836 this_translated = 0; 1844 this_translated = 0;
1845 ch = -1;
1837 } 1846 }
1838 if (ch > 0400) 1847
1839 j = ((unsigned char) ch | 0200); 1848 /* BYTE = last byte of character CH when represented as text */
1840 else 1849 j = byte;
1841 j = (unsigned char) ch;
1842 1850
1843 if (i == infinity) 1851 if (i == infinity)
1844 stride_for_teases = BM_tab[j]; 1852 stride_for_teases = BM_tab[j];
1845 BM_tab[j] = dirlen - i; 1853 BM_tab[j] = dirlen - i;
1846 /* A translation table is accompanied by its inverse -- see 1854 /* A translation table is accompanied by its inverse -- see
1847 comment in casetab.c. */ 1855 comment in casetab.c. */
1848 if (this_translated) 1856 if (this_translated)
1849 { 1857 {
1850 Ichar starting_ch = ch; 1858 Ichar starting_ch = ch;
1851 EMACS_INT starting_j = j; 1859 EMACS_INT starting_j = j;
1860
1861 text_checking_assert (valid_ichar_p (ch));
1852 do 1862 do
1853 { 1863 {
1854 ch = TRANSLATE (inverse_trt, ch); 1864 ch = TRANSLATE (inverse_trt, ch);
1855 1865
1856 if (ch > 0x7F && buffer_entirely_one_byte_p) 1866 if (ch > 0x7F && buffer_entirely_one_byte_p)
1857 continue; 1867 continue;
1858 1868
1859 if (ch > 0xFF && buffer_nothing_greater_than_0xff) 1869 if (ch > 0xFF && buffer_nothing_greater_than_0xff)
1860 continue; 1870 continue;
1861 1871
1862 if (ch > 0400) 1872
1863 j = ((unsigned char) ch | 0200); 1873 /* Retrieve last byte of character CH when represented as
1864 else 1874 text */
1865 j = (unsigned char) ch; 1875 {
1866 1876 Ibyte tmp[MAX_ICHAR_LEN];
1877 Bytecount chlen;
1878
1879 chlen = set_itext_ichar (tmp, ch);
1880 j = tmp[chlen - 1];
1881 }
1882
1867 /* For all the characters that map into CH, set up 1883 /* For all the characters that map into CH, set up
1868 simple_translate to map the last byte into 1884 simple_translate to map the last byte into
1869 STARTING_J. */ 1885 STARTING_J. */
1870 simple_translate[j] = (Ibyte) starting_j; 1886 simple_translate[j] = (Ibyte) starting_j;
1871 BM_tab[j] = dirlen - i; 1887 BM_tab[j] = dirlen - i;
1872 1888
1873 } while (ch != starting_ch); 1889 }
1890 while (ch != starting_ch);
1874 } 1891 }
1875 #else 1892 #else /* not MULE */
1876 EMACS_INT k; 1893 EMACS_INT k;
1877 j = *ptr; 1894 j = *ptr;
1878 k = (j = TRANSLATE (trt, j)); 1895 k = (j = TRANSLATE (trt, j));
1879 if (i == infinity) 1896 if (i == infinity)
1880 stride_for_teases = BM_tab[j]; 1897 stride_for_teases = BM_tab[j];
1884 while ((j = TRANSLATE (inverse_trt, j)) != k) 1901 while ((j = TRANSLATE (inverse_trt, j)) != k)
1885 { 1902 {
1886 simple_translate[j] = (Ibyte) k; 1903 simple_translate[j] = (Ibyte) k;
1887 BM_tab[j] = dirlen - i; 1904 BM_tab[j] = dirlen - i;
1888 } 1905 }
1889 #endif 1906 #endif /* (not) MULE */
1890 } 1907 }
1891 else 1908 else
1892 { 1909 {
1893 j = *ptr; 1910 j = *ptr;
1894 1911