- Inkscape Devel - Inkscape Mailman

printing + gradient
by herve couvelard 13 Feb '04

13 Feb '04

bulia byak wrote: > My guess is that you print via Postscript, and Postscript does not > support gradients. When other vector apps export things like > transparency or gradients to PS, they "simulate" them, usually by > bitmaps. Inkscape does not yet do that. But it's just a guess, your bug > may be different. Please do file a bug report. well.. just a 2 cents (euro) stupid idea/question i export from inkscape to png a gradient i import that png into ooffice i print the document (via cups on redhat )? postcript ? the gradient print well... ooffice did the bitmap conversion ? cups don't use postcript ? i understood nothing ? Hervé

1 0

bug with moving selections
by John Cliff 13 Feb '04

13 Feb '04

Hi folks, Found a bug that seems a bit weird, If you create a shape, doesnt matter what, go to select and drag it, but move the mouse outside of the drawing area, it becomes impossible to interact with the interface again until you do another select action. Ie if you try to change tool, use a menu, close the program etc, it behaves like its hung. If the shape is the only thing you've got in the document the only way to get control back is to click on it again, made me think the whole thing had died for a minute there. Have only tried it on Win 2k, but it does it in both the builds I've got on my home box, and the 2 on my work one. cheers John btw isnt bug 894662 closed? __________________________________ Do you Yahoo!? Yahoo! Finance: Get your refund fast by filing online. http://taxes.yahoo.com/filing.html

2 1

RE: [Inkscape-devel] bug with moving selections
by bulia byak 13 Feb '04

13 Feb '04

> Found a bug that seems a bit weird, >If you create a shape, doesnt matter what, go to >select and drag it, but move the mouse >outside of the drawing area, it becomes impossible to >interact with the interface again until you do another >select action. Ie if you try to change tool, use a >menu, close the program etc, it behaves like its hung. >If the shape is the only thing you've got in the >document the only way to get control back is to click >on it again, made me think the whole thing had died >for a minute there. That seems to be http://sourceforge.net/tracker/index.php?func=detail&aid=852976&group_id=93… Obviously the reason is that on Windows, GTK sends different (or no) events when mouse is outside the window. This needs investigation, unfortunately we have only one developer actively working on the Windows version. >btw isnt bug 894662 closed? I guess it waits for confirmation. If you can confirm that you can import PNG, GIF, JPEG with the latest win32 build, I'll close it. _________________________________________________________________ Help STOP SPAM with the new MSN 8 and get 2 months FREE* http://join.msn.com/?page=dept/bcomm&pgmarket=en-ca&RU=http%3a%2f%2fjoin.ms…

2 1

RE: [Inkscape-devel] fr.po (again)
by bulia byak 13 Feb '04

13 Feb '04

>I join an update fr.po . Merci beacoup. It's uploaded to CVS. Now, if you can compile the latest CVS with French interface, it would be nice if you could send us a screenshot (with some dialogs, menus etc.) for the site. _________________________________________________________________ MSN 8 helps eliminate e-mail viruses. Get 2 months FREE*. http://join.msn.com/?page=features/virus&pgmarket=en-ca&RU=http%3a%2f%2fjoi…

3 2

ru.po
by Alexey Remizov 13 Feb '04

13 Feb '04

Hello. Here is russian messages based on inkscape.pot of 0.37 release and messages from sodipodi 0.33 translation. 535 translated messages, 8 fuzzy translations, 57 untranslated messages. -- С уважением. | WBR. | Алексей. | Alexey. | ----------------------------| mailto:alexey@...188... | jabber:remizov@...189... | ----------------------------'

1 0

PROPOSAL: Reorganizing Directories
by Bryce Harrington 13 Feb '04

13 Feb '04

I'd like to propose some structural changes to both the codebase directory hierarchy and the directories created during installation. If these changes are acceptable, they could be introduced for the 0.38 development cycle. Bryce Distribution / Packaging Files ============================== Presently, we have a number of files in the codebase related to generation of distribution packages. Some of these files are in the root inkscape/ directory, others are in subdirs like inkscape/debian/. To head off potential clutter as Inkscape becomes supported on a wider variety of systems, we should collect as many of these packaging files as feasible into a subdir called inkscape/packaging/. Inside that directory would be distro-specific subdirs, plus a common directory for shared scripts and collateral: inkscape/packaging/ common/ debian/ fedora/ fink/ mandrake/ suse/ Config Files ============ Currently we do not have global config files, but instead put the config file in the user's home dir in ~/.inkscape/. These are generated at run-time by Inkscape itself, so the defaults are hard-coded into the application, and can't be customized by the sysadmin except by editing the appropriate header file and recompiling/repackaging the app. The correct approach for Linux apps is to have both user and global configs, and to use the global as the default if the user does not yet have a config file. This way, site customizations are known (such as the fact that its shared files are in /usr/share, vs. /usr/local/share, vs. /opt/inkscape/share, etc.) The proposal is for Inkscape to establish its config file location in /etc/inkscape/, and to place a global preferences.xml file there, as well as any other config files we need later, like extensions.xml, etc. Ideally, the global config file would be loaded first, and then the user config file "override" any portion of the global options. This way the user config file only needs to contain customizations, so that if we choose to change the defaults in a future install, the user's config files do not necessarily become invalid. Inside the CVS codebase, the default etc. files would be stored in inkscape/etc/, instead of being produced by the application at runtime. During installation, installation-specific parameters (such as alternate paths for various files) will be substituted in appropriately. "Share" Collateral ================== Inkscape currently generates a /usr/local/share/inkscape/ directory during install, and copies the contents of the codebase's icons/ directory to that location. We need more pieces of collateral installed, and need to do so in a more organized manner. The recommendation is to establish a set of subdirs in /usr/local/share/inkscape/, as follows: AUTHORS NEWS clipart/ examples/ extensions/ fonts/ gradients/ icons/ keyboards/ markers/ palettes/ patterns/ screens/ about.svg templates/ tutorials/ This uses a flat hierarchy for the sake of simplicity. In the CVS codebase, all of these would be placed in inkscape/share/. This way, the installation script could be set up to recursively copy everything in that directory to the target share directory, and we won't need to specify each of the subdirs in that script, which will make maintenance less hassle. For clipart, I think we definitely want to consider being able to tie in with "external" clipart collections. One way would be to simply put symlinks in the clipart dir to those collections. This way, if the flags package were to install into /usr/share/flags-svg/, we'd just symlink there. Possibly similarly for extensions. I think an argument could be made that the tutorials ought to be placed somewhere under the /usr/doc/ or /usr/local/doc/ path, so I'm not 100% certain of having the tutorials located here.

3 4

openoffice fileformat
by Johan Ceuppens 13 Feb '04

13 Feb '04

Hi, I've tested libinkjar and it can read openoffice fileformats now. I'd like to know if I can hook this up and move it to cvs HEAD. I was thinking of bringing up a GUI which greps/seeks the jar file for filenames and letting the user choose which SVGs to open. If you'd like different or additional behaviour, please share with the list. Johan

3 2

Formatting code
by Bob Jamison 13 Feb '04

13 Feb '04

Re: our conversation on Jabber about formatting data for the repr tree, here is some code I appropriated from Apache's APR lib, which can give us total control over formatting output strings. With Mental and Bryce's work in this area, these files might not be necessary, but then again they might be useful, so here they are. One very nice function for the repr tree would be: sp_repr_printf(repr, key, "format", args...) Would relieve the need for a lot of temporary buffers all over the place. Bob /*######################################################################## ## $Id$ ########################################################################*/ /* * Print formatter. This file's purpose is to * allow Inkscape developers to have complete * control over formatted strings. This can be * tweaked to suit. * * This file was basically copied from Apache's * apr_snprintf.c, but with extensive modifications * to ease embedding in a GLib project. Their license * is below. * * Copyright (C) 1999-2003 Authors * * Released under GNU GPL, read the file 'COPYING' for more information * */ /* ==================================================================== * The Apache Software License, Version 1.1 * * Copyright (c) 2000-2003 The Apache Software Foundation. All rights * reserved. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in * the documentation and/or other materials provided with the * distribution. * * 3. The end-user documentation included with the redistribution, * if any, must include the following acknowledgment: * "This product includes software developed by the * Apache Software Foundation (http://www.apache.org/)." * Alternately, this acknowledgment may appear in the software itself, * if and wherever such third-party acknowledgments normally appear. * * 4. The names "Apache" and "Apache Software Foundation" must * not be used to endorse or promote products derived from this * software without prior written permission. For written * permission, please contact apache@...199... * * 5. Products derived from this software may not be called "Apache", * nor may "Apache" appear in their name, without prior written * permission of the Apache Software Foundation. * * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESSED OR IMPLIED * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE * DISCLAIMED. IN NO EVENT SHALL THE APACHE SOFTWARE FOUNDATION OR * ITS CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF * USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, * OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT * OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF * SUCH DAMAGE. * ==================================================================== * * This software consists of voluntary contributions made by many * individuals on behalf of the Apache Software Foundation. For more * information on the Apache Software Foundation, please see * <http://www.apache.org/>. * * The ink_vsnprintf/ink_snprintf functions are based on, and used with the * permission of, the SIO stdio-replacement strx_* functions by Panos * Tsirigotis <panos@...200...> for xinetd. */ #include <math.h> #include <ctype.h> #include <string.h> #include <glib.h> #include "ink-snprintf.h" /*#################################### # D E F I N I T I O N S ####################################*/ /** * Structure used by the variable-formatter routines. */ struct ink_vformatter_buff_t { gchar *curpos;/** The current position */ gchar *endpos;/** The end position of the format string */ }; typedef struct ink_vformatter_buff_t ink_vformatter_buff_t; typedef enum { NO = 0, YES = 1 } boolean_e; #ifndef FALSE #define FALSE 0 #endif #ifndef TRUE #define TRUE 1 #endif #define NUL '\0' #define S_NULL "(null)" #define S_NULL_LEN 6 #define FLOAT_DIGITS 6 #define EXPONENT_LENGTH 10 /* * NUM_BUF_SIZE is the size of the buffer used for arithmetic conversions * * NOTICE: this is a magic number; do not decrease it */ #define NUM_BUF_SIZE 512 /* * cvt.c - IEEE floating point formatting routines for FreeBSD * from GNU libc-4.6.27. Modified to be thread safe. */ /* * ink_ecvt converts to decimal * the number of digits is specified by ndigit * decpt is set to the position of the decimal point * sign is set to 0 for positive, 1 for negative */ #define NDIG 80 /*#################################### # M A C R O S ####################################*/ /* * The INS_CHAR macro inserts a character in the buffer and writes * the buffer back to disk if necessary * It uses the char pointers sp and bep: * sp points to the next available character in the buffer * bep points to the end-of-buffer+1 * While using this macro, note that the nextb pointer is NOT updated. * * NOTE: Evaluation of the c argument should not have any side-effects */ #define INS_CHAR(c, sp, bep, cc) \ { \ if (sp) { \ if (sp >= bep) { \ vbuff->curpos = sp; \ if (flush_func(vbuff)) \ return -1; \ sp = vbuff->curpos; \ bep = vbuff->endpos; \ } \ *sp++ = (c); \ } \ cc++; \ } #define NUM(c) (c - '0') #define STR_TO_DEC(str, num) \ num = NUM(*str++); \ while (g_ascii_isdigit(*str)) \ { \ num *= 10 ; \ num += NUM(*str++); \ } /* * This macro does zero padding so that the precision * requirement is satisfied. The padding is done by * adding '0's to the left of the string that is going * to be printed. We don't allow precision to be large * enough that we continue past the start of s. * * NOTE: this makes use of the magic info that s is * always based on num_buf with a size of NUM_BUF_SIZE. */ #define FIX_PRECISION(adjust, precision, s, s_len) \ if (adjust) { \ gint p = precision < NUM_BUF_SIZE - 1 ? \ precision : NUM_BUF_SIZE - 1; \ while (s_len < p) { \ *--s = '0'; \ s_len++; \ } \ } /* * Macro that does padding. The padding is done by printing * the character ch. */ #define PAD(width, len, ch) \ do \ { \ INS_CHAR(ch, sp, bep, cc); \ width--; \ } \ while (width > len) /* * Prefix the character ch to the string str * Increase length * Set the has_prefix flag */ #define PREFIX(str, length, ch) \ *--str = ch; \ length++; \ has_prefix=YES; /*#################################### # CVT - float-to-ascii conversion ####################################*/ /* buf must have at least NDIG bytes */ static gchar * ink_cvt(gdouble arg, gint ndigits, gint *decpt, gint *sign, gint eflag, gchar *buf) { gint r2; gdouble fi, fj; gchar *p, *p1; if (ndigits >= NDIG - 1) ndigits = NDIG - 2; r2 = 0; *sign = 0; p = &buf[0]; if (arg < 0) { *sign = 1; arg = -arg; } arg = modf(arg, &fi); p1 = &buf[NDIG]; /* * Do integer part */ if (fi != 0) { p1 = &buf[NDIG]; while (p1 > &buf[0] && fi != 0) { fj = modf(fi / 10, &fi); *--p1 = (int) ((fj + .03) * 10) + '0'; r2++; } while (p1 < &buf[NDIG]) *p++ = *p1++; } else if (arg > 0) { while ((fj = arg * 10) < 1) { arg = fj; r2--; } } p1 = &buf[ndigits]; if (eflag == 0) p1 += r2; *decpt = r2; if (p1 < &buf[0]) { buf[0] = '\0'; return (buf); } while (p <= p1 && p < &buf[NDIG]) { arg *= 10; arg = modf(arg, &fj); *p++ = (int) fj + '0'; } if (p1 >= &buf[NDIG]) { buf[NDIG - 1] = '\0'; return (buf); } p = p1; *p1 += 5; while (*p1 > '9') { *p1 = '0'; if (p1 > buf) ++ * --p1; else { *p1 = '1'; (*decpt)++; if (eflag == 0) { if (p > buf) *p = '0'; p++; } } } *p = '\0'; return (buf); } static gchar * ink_ecvt(gdouble arg, gint ndigits, gint *decpt, gint *sign, gchar *buf) { return (ink_cvt(arg, ndigits, decpt, sign, 1, buf)); } static gchar * ink_fcvt(gdouble arg, gint ndigits, gint *decpt, gint *sign, gchar *buf) { return (ink_cvt(arg, ndigits, decpt, sign, 0, buf)); } /* * ink_gcvt - Floating output conversion to * minimal length string */ static gchar * ink_gcvt(gdouble number, gint ndigit, gchar *buf, boolean_e altform) { gint sign, decpt; gchar *p1, *p2; gint i; gchar buf1[NDIG]; p1 = ink_ecvt(number, ndigit, &decpt, &sign, buf1); p2 = buf; if (sign) *p2++ = '-'; for (i = ndigit - 1; i > 0 && p1[i] == '0'; i--) ndigit--; if ((decpt >= 0 && decpt - ndigit > 4) || (decpt < 0 && decpt < -3)) { /* use E-style */ decpt--; *p2++ = *p1++; *p2++ = '.'; for (i = 1; i < ndigit; i++) *p2++ = *p1++; *p2++ = 'e'; if (decpt < 0) { decpt = -decpt; *p2++ = '-'; } else *p2++ = '+'; if (decpt / 100 > 0) *p2++ = decpt / 100 + '0'; if (decpt / 10 > 0) *p2++ = (decpt % 100) / 10 + '0'; *p2++ = decpt % 10 + '0'; } else { if (decpt <= 0) { if (*p1 != '0') *p2++ = '.'; while (decpt < 0) { decpt++; *p2++ = '0'; } } for (i = 1; i <= ndigit; i++) { *p2++ = *p1++; if (i == decpt) *p2++ = '.'; } if (ndigit < decpt) { while (ndigit++ < decpt) *p2++ = '0'; *p2++ = '.'; } } if (p2[-1] == '.' && !altform) p2--; *p2 = '\0'; return (buf); } /* * Convert num to its decimal format. * Return value: * - a pointer to a string containing the number (no sign) * - len contains the length of the string * - is_negative is set to TRUE or FALSE depending on the sign * of the number (always set to FALSE if is_unsigned is TRUE) * * The caller provides a buffer for the string: that is the buf_end argument * which is a pointer to the END of the buffer + 1 (i.e. if the buffer * is declared as buf[ 100 ], buf_end should be &buf[ 100 ]) * * Note: we have 2 versions. One is used when we need to use quads * (conv_10_quad), the other when we don't (conv_10). We're assuming the * latter is faster. */ static gchar * conv_10(gint32 num, gboolean is_unsigned, gboolean *is_negative, gchar *buf_end, gint *len) { gchar *p = buf_end; guint32 magnitude; if (is_unsigned) { magnitude = (guint32) num; *is_negative = FALSE; } else { *is_negative = (num < 0); /* * On a 2's complement machine, negating the most negative integer * results in a number that cannot be represented as a signed integer. * Here is what we do to obtain the number's magnitude: * a. add 1 to the number * b. negate it (becomes positive) * c. convert it to unsigned * d. add 1 */ if (*is_negative) { gint32 t = num + 1; magnitude = ((guint32) -t) + 1; } else magnitude = (guint32) num; } /* * We use a do-while loop so that we write at least 1 digit */ do { guint32 new_magnitude = magnitude / 10; *--p = (gchar) (magnitude - new_magnitude * 10 + '0'); magnitude = new_magnitude; } while (magnitude); *len = buf_end - p; return (p); } static gchar * conv_10_quad(gint64 num, gboolean is_unsigned, gboolean *is_negative, gchar *buf_end, gint *len) { gchar *p = buf_end; guint64 magnitude; /* * We see if we can use the faster non-quad version by checking the * number against the largest long value it can be. If <=, we * punt to the quicker version. */ if ( (num <= ULONG_MAX && is_unsigned) || (num <= LONG_MAX && !is_unsigned) ) return(conv_10( (gint32)num, is_unsigned, is_negative, buf_end, len)); if (is_unsigned) { magnitude = (guint64) num; *is_negative = FALSE; } else { *is_negative = (num < 0); /* * On a 2's complement machine, negating the most negative integer * results in a number that cannot be represented as a signed integer. * Here is what we do to obtain the number's magnitude: * a. add 1 to the number * b. negate it (becomes positive) * c. convert it to unsigned * d. add 1 */ if (*is_negative) { gint64 t = num + 1; magnitude = ((guint64) -t) + 1; } else magnitude = (guint64) num; } /* * We use a do-while loop so that we write at least 1 digit */ do { guint64 new_magnitude = magnitude / 10; *--p = (gchar) (magnitude - new_magnitude * 10 + '0'); magnitude = new_magnitude; } while (magnitude); *len = buf_end - p; return (p); } /* * Convert a floating point number to a string formats 'f', 'e' or 'E'. * The result is placed in buf, and len denotes the length of the string * The sign is returned in the is_negative argument (and is not placed * in buf). */ static gchar * conv_fp(gchar format, gdouble num, boolean_e add_dp, gint precision, gboolean *is_negative, gchar *buf, gint *len) { gchar *s = buf; gchar *p; gint decimal_point; gchar buf1[NDIG]; if (format == 'f') p = ink_fcvt(num, precision, &decimal_point, is_negative, buf1); else /* either e or E format */ p = ink_ecvt(num, precision + 1, &decimal_point, is_negative, buf1); /* * Check for Infinity and NaN */ if (g_ascii_isalpha(*p)) { *len = strlen(p); memcpy(buf, p, *len + 1); *is_negative = FALSE; return (buf); } if (format == 'f') { if (decimal_point <= 0) { *s++ = '0'; if (precision > 0) { *s++ = '.'; while (decimal_point++ < 0) *s++ = '0'; } else if (add_dp) *s++ = '.'; } else { while (decimal_point-- > 0) *s++ = *p++; if (precision > 0 || add_dp) *s++ = '.'; } } else { *s++ = *p++; if (precision > 0 || add_dp) *s++ = '.'; } /* * copy the rest of p, the NUL is NOT copied */ while (*p) *s++ = *p++; if (format != 'f') { gchar temp[EXPONENT_LENGTH]; /* for exponent conversion */ gint t_len; gboolean exponent_is_negative; *s++ = format; /* either e or E */ decimal_point--; if (decimal_point != 0) { p = conv_10((gint32) decimal_point, FALSE, &exponent_is_negative, &temp[EXPONENT_LENGTH], &t_len); *s++ = exponent_is_negative ? '-' : '+'; /* * Make sure the exponent has at least 2 digits */ if (t_len == 1) *s++ = '0'; while (t_len--) *s++ = *p++; } else { *s++ = '+'; *s++ = '0'; *s++ = '0'; } } *len = s - buf; return (buf); } /* * Convert num to a base X number where X is a power of 2. nbits determines X. * For example, if nbits is 3, we do base 8 conversion * Return value: * a pointer to a string containing the number * * The caller provides a buffer for the string: that is the buf_end argument * which is a pointer to the END of the buffer + 1 (i.e. if the buffer * is declared as buf[ 100 ], buf_end should be &buf[ 100 ]) * * As with conv_10, we have a faster version which is used when * the number isn't quad size. */ static gchar * conv_p2(guint32 num, gint nbits, gchar format, gchar *buf_end, gint *len) { gint mask = (1 << nbits) - 1; gchar *p = buf_end; const gchar low_digits[] = "0123456789abcdef"; const gchar upper_digits[] = "0123456789ABCDEF"; const gchar *digits = (format == 'X') ? upper_digits : low_digits; do { *--p = digits[num & mask]; num >>= nbits; } while (num); *len = buf_end - p; return (p); } static gchar * conv_p2_quad(guint64 num, gint nbits, gchar format, gchar *buf_end, gint *len) { gint mask = (1 << nbits) - 1; gchar *p = buf_end; const gchar low_digits[] = "0123456789abcdef"; const gchar upper_digits[] = "0123456789ABCDEF"; const gchar *digits = (format == 'X') ? upper_digits : low_digits; if (num <= ULONG_MAX) return(conv_p2((guint32)num, nbits, format, buf_end, len)); do { *--p = digits[num & mask]; num >>= nbits; } while (num); *len = buf_end - p; return (p); } /*#################################### # F O R M A T T I N G ####################################*/ /* * Do format conversion placing the output in buffer */ static gint ink_vformatter(gint (*flush_func)(ink_vformatter_buff_t *), ink_vformatter_buff_t *vbuff, const gchar *fmt, va_list ap) { gchar *sp; gchar *bep; gint cc = 0; gint i; gchar *s = NULL; gchar *q; gint s_len; gint min_width = 0; gint precision = 0; enum { LEFT, RIGHT } adjust; gchar pad_char; gchar prefix_char; gdouble fp_num; gint64 i_quad = (gint64) 0; guint64 ui_quad; gint32 i_num = (gint32) 0; guint32 ui_num; gchar num_buf[NUM_BUF_SIZE]; gchar char_buf[2]; /* for printing %% and %<unknown> */ enum var_type_enum { IS_QUAD, IS_LONG, IS_SHORT, IS_INT }; enum var_type_enum var_type = IS_INT; /* * Flag variables */ boolean_e alternate_form; boolean_e print_sign; boolean_e print_blank; boolean_e adjust_precision; boolean_e adjust_width; gboolean is_negative; sp = vbuff->curpos; bep = vbuff->endpos; while (*fmt) { if (*fmt != '%') { INS_CHAR(*fmt, sp, bep, cc); } else { /* * Default variable settings */ boolean_e print_something = YES; adjust = RIGHT; alternate_form = print_sign = print_blank = NO; pad_char = ' '; prefix_char = NUL; fmt++; /* * Try to avoid checking for flags, width or precision */ if (!g_ascii_islower(*fmt)) { /* * Recognize flags: -, #, BLANK, + */ for (;; fmt++) { if (*fmt == '-') adjust = LEFT; else if (*fmt == '+') print_sign = YES; else if (*fmt == '#') alternate_form = YES; else if (*fmt == ' ') print_blank = YES; else if (*fmt == '0') pad_char = '0'; else break; } /* * Check if a width was specified */ if (g_ascii_isdigit(*fmt)) { STR_TO_DEC(fmt, min_width); adjust_width = YES; } else if (*fmt == '*') { min_width = va_arg(ap, int); fmt++; adjust_width = YES; if (min_width < 0) { adjust = LEFT; min_width = -min_width; } } else adjust_width = NO; /* * Check if a precision was specified */ if (*fmt == '.') { adjust_precision = YES; fmt++; if ( g_ascii_isdigit (*fmt)) { STR_TO_DEC(fmt, precision); } else if (*fmt == '*') { precision = va_arg(ap, int); fmt++; if (precision < 0) precision = 0; } else precision = 0; } else adjust_precision = NO; } else adjust_precision = adjust_width = NO; /* * Modifier check */ if (*fmt == 'q') { var_type = IS_QUAD; fmt++; } else if (*fmt == 'l') { var_type = IS_LONG; fmt++; } else if (*fmt == 'h') { var_type = IS_SHORT; fmt++; } else { var_type = IS_INT; } /* * Argument extraction and printing. * First we determine the argument type. * Then, we convert the argument to a string. * On exit from the switch, s points to the string that * must be printed, s_len has the length of the string * The precision requirements, if any, are reflected in s_len. * * NOTE: pad_char may be set to '0' because of the 0 flag. * It is reset to ' ' by non-numeric formats */ switch (*fmt) { case 'u': if (var_type == IS_QUAD) { i_quad = va_arg(ap, guint64); s = conv_10_quad(i_quad, 1, &is_negative, &num_buf[NUM_BUF_SIZE], &s_len); } else { if (var_type == IS_LONG) i_num = (gint32) va_arg(ap, guint32); else if (var_type == IS_SHORT) i_num = (gint32) (unsigned short) va_arg(ap, unsigned int); else i_num = (gint32) va_arg(ap, unsigned int); s = conv_10(i_num, 1, &is_negative, &num_buf[NUM_BUF_SIZE], &s_len); } FIX_PRECISION(adjust_precision, precision, s, s_len); break; case 'd': case 'i': if (var_type == IS_QUAD) { i_quad = va_arg(ap, gint64); s = conv_10_quad(i_quad, 0, &is_negative, &num_buf[NUM_BUF_SIZE], &s_len); } else { if (var_type == IS_LONG) i_num = (gint32) va_arg(ap, gint32); else if (var_type == IS_SHORT) i_num = (gint32) (short) va_arg(ap, int); else i_num = (gint32) va_arg(ap, int); s = conv_10(i_num, 0, &is_negative, &num_buf[NUM_BUF_SIZE], &s_len); } FIX_PRECISION(adjust_precision, precision, s, s_len); if (is_negative) prefix_char = '-'; else if (print_sign) prefix_char = '+'; else if (print_blank) prefix_char = ' '; break; case 'o': if (var_type == IS_QUAD) { ui_quad = va_arg(ap, guint64); s = conv_p2_quad(ui_quad, 3, *fmt, &num_buf[NUM_BUF_SIZE], &s_len); } else { if (var_type == IS_LONG) ui_num = (guint32) va_arg(ap, guint32); else if (var_type == IS_SHORT) ui_num = (guint32) (unsigned short) va_arg(ap, unsigned int); else ui_num = (guint32) va_arg(ap, unsigned int); s = conv_p2(ui_num, 3, *fmt, &num_buf[NUM_BUF_SIZE], &s_len); } FIX_PRECISION(adjust_precision, precision, s, s_len); if (alternate_form && *s != '0') { *--s = '0'; s_len++; } break; case 'x': case 'X': if (var_type == IS_QUAD) { ui_quad = va_arg(ap, guint64); s = conv_p2_quad(ui_quad, 4, *fmt, &num_buf[NUM_BUF_SIZE], &s_len); } else { if (var_type == IS_LONG) ui_num = (guint32) va_arg(ap, guint32); else if (var_type == IS_SHORT) ui_num = (guint32) (unsigned short) va_arg(ap, unsigned int); else ui_num = (guint32) va_arg(ap, unsigned int); s = conv_p2(ui_num, 4, *fmt, &num_buf[NUM_BUF_SIZE], &s_len); } FIX_PRECISION(adjust_precision, precision, s, s_len); if (alternate_form && i_num != 0) { *--s = *fmt; /* 'x' or 'X' */ *--s = '0'; s_len += 2; } break; case 's': s = va_arg(ap, gchar *); if (s != NULL) { if (!adjust_precision) { s_len = strlen(s); } else { /* From the C library standard in section 7.9.6.1: * ...if the precision is specified, no more then * that many characters are written. If the * precision is not specified or is greater * than the size of the array, the array shall * contain a null character. * * My reading is is precision is specified and * is less then or equal to the size of the * array, no null character is required. So * we can't do a strlen. * * This figures out the length of the string * up to the precision. Once it's long enough * for the specified precision, we don't care * anymore. * * NOTE: you must do the length comparison * before the check for the null character. * Otherwise, you'll check one beyond the * last valid character. */ const gchar *walk; for (walk = s, s_len = 0; (s_len < precision) && (*walk != '\0'); ++walk, ++s_len); } } else { s = S_NULL; s_len = S_NULL_LEN; } pad_char = ' '; break; case 'f': case 'e': case 'E': fp_num = va_arg(ap, double); /* * We use &num_buf[ 1 ], so that we have room for the sign */ s = NULL; #ifdef HAVE_ISNAN if (isnan(fp_num)) { s = "nan"; s_len = 3; } #endif #ifdef HAVE_ISINF if (!s && isinf(fp_num)) { s = "inf"; s_len = 3; } #endif if (!s) { s = conv_fp(*fmt, fp_num, alternate_form, (adjust_precision == NO) ? FLOAT_DIGITS : precision, &is_negative, &num_buf[1], &s_len); if (is_negative) prefix_char = '-'; else if (print_sign) prefix_char = '+'; else if (print_blank) prefix_char = ' '; } break; case 'g': case 'G': if (adjust_precision == NO) precision = FLOAT_DIGITS; else if (precision == 0) precision = 1; /* * * We use &num_buf[ 1 ], so that we have room for the sign */ s = ink_gcvt(va_arg(ap, double), precision, &num_buf[1], alternate_form); if (*s == '-') prefix_char = *s++; else if (print_sign) prefix_char = '+'; else if (print_blank) prefix_char = ' '; s_len = strlen(s); if (alternate_form && (q = strchr(s, '.')) == NULL) { s[s_len++] = '.'; s[s_len] = '\0'; /* delimit for following strchr() */ } if (*fmt == 'G' && (q = strchr(s, 'e')) != NULL) *q = 'E'; break; case 'c': char_buf[0] = (char) (va_arg(ap, int)); s = &char_buf[0]; s_len = 1; pad_char = ' '; break; case '%': char_buf[0] = '%'; s = &char_buf[0]; s_len = 1; pad_char = ' '; break; case 'n': if (var_type == IS_QUAD) *(va_arg(ap, gint64 *)) = cc; else if (var_type == IS_LONG) *(va_arg(ap, long *)) = cc; else if (var_type == IS_SHORT) *(va_arg(ap, short *)) = cc; else *(va_arg(ap, int *)) = cc; print_something = NO; break; /* * This is where we extend the printf format, with a second * type specifier */ case 'p': switch(*++fmt) { /* * If the pointer size is equal to or smaller than the size * of the largest unsigned int, we convert the pointer to a * hex number, otherwise we print "%p" to indicate that we * don't handle "%p". */ case 'p': #ifdef VOID_P_IS_QUAD if (sizeof(void *) <= sizeof(guint64)) { ui_quad = (guint64) va_arg(ap, void *); s = conv_p2_quad(ui_quad, 4, 'x', &num_buf[NUM_BUF_SIZE], &s_len); } #else if (sizeof(void *) <= sizeof(guint32)) { ui_num = (guint32) va_arg(ap, void *); s = conv_p2(ui_num, 4, 'x', &num_buf[NUM_BUF_SIZE], &s_len); } #endif else { s = "%p"; s_len = 2; prefix_char = NUL; } pad_char = ' '; break; case 'T': char_buf[0] = '0'; s = &char_buf[0]; s_len = 1; pad_char = ' '; break; case NUL: /* if %p ends the string, oh well ignore it */ continue; default: s = "bogus %p"; s_len = 8; prefix_char = NUL; (void)va_arg(ap, void *); /* skip the bogus argument on the stack */ break; } break; case NUL: /* * The last character of the format string was %. * We ignore it. */ continue; /* * The default case is for unrecognized %'s. * We print %<char> to help the user identify what * option is not understood. * This is also useful in case the user wants to pass * the output of format_converter to another function * that understands some other %<char> (like syslog). * Note that we can't point s inside fmt because the * unknown <char> could be preceded by width etc. */ default: char_buf[0] = '%'; char_buf[1] = *fmt; s = char_buf; s_len = 2; pad_char = ' '; break; } if (prefix_char != NUL && s != S_NULL && s != char_buf) { *--s = prefix_char; s_len++; } if (adjust_width && adjust == RIGHT && min_width > s_len) { if (pad_char == '0' && prefix_char != NUL) { INS_CHAR(*s, sp, bep, cc); s++; s_len--; min_width--; } PAD(min_width, s_len, pad_char); } /* * Print the string s. */ if (print_something == YES) { for (i = s_len; i != 0; i--) { INS_CHAR(*s, sp, bep, cc); s++; } } if (adjust_width && adjust == LEFT && min_width > s_len) PAD(min_width, s_len, pad_char); } fmt++; } vbuff->curpos = sp; return cc; } static int snprintf_flush(ink_vformatter_buff_t *vbuff) { /* if the buffer fills we have to abort immediately, there is no way * to "flush" an ink_snprintf... there's nowhere to flush it to. */ return -1; } gint ink_snprintf(gchar *buf, gint len, const gchar *format, ...) { gint cc; va_list ap; ink_vformatter_buff_t vbuff; if (len == 0) { /* NOTE: This is a special case; we just want to return the number * of chars that would be written (minus \0) if the buffer * size was infinite. We leverage the fact that INS_CHAR * just does actual inserts iff the buffer pointer is non-NULL. * In this case, we don't care what buf is; it can be NULL, since * we don't touch it at all. */ vbuff.curpos = NULL; vbuff.endpos = NULL; } else { /* save one byte for nul terminator */ vbuff.curpos = buf; vbuff.endpos = buf + len - 1; } va_start(ap, format); cc = ink_vformatter(snprintf_flush, &vbuff, format, ap); va_end(ap); if (len != 0) { *vbuff.curpos = '\0'; } return (cc == -1) ? (gint)len : cc; } gint ink_vsnprintf(gchar *buf, gint len, const gchar *format, va_list ap) { gint cc; ink_vformatter_buff_t vbuff; if (len == 0) { /* See above note */ vbuff.curpos = NULL; vbuff.endpos = NULL; } else { /* save one byte for nul terminator */ vbuff.curpos = buf; vbuff.endpos = buf + len - 1; } cc = ink_vformatter(snprintf_flush, &vbuff, format, ap); if (len != 0) { *vbuff.curpos = '\0'; } return (cc == -1) ? (gint)len : cc; } #ifndef __INK_SNPRINTF_H__ #define __INK_SNPRINTF_H__ /*######################################################################## ## $Id$ ########################################################################*/ /* * Print formatter. This file's purpose is to * allow Inkscape developers to have complete * control over formatted strings. This can be * tweaked to suit. * * Copyright (C) 1999-2003 Authors * * Released under GNU GPL, read the file 'COPYING' for more information */ #include <glib.h> #include <stdarg.h> /** * Create a formatted string with a given format string * and a standard variable argument list. */ gint ink_snprintf(gchar *buf, gint len, const gchar *format, ...); /** * Create a formatted string with a given format string * and argument list. */ gint ink_vsnprintf(gchar *buf, gint len, const gchar *format, va_list ap); #endif /* __INK_SNPRINTF_H__ */

1 0

Re: [Inkscape-devel] Some fairly random feature reqs + bug reports.
by bulia byak 13 Feb '04

13 Feb '04

>>>What happened in the Fill/stroke dialog? the entry >>>boxs for RGB values on the wheel tab are gone >> >> >>That's temporary, we plan to restore them, although not RGB - the wheel is >>HSV and it will have HSV fields, if you need RGB switch to the RGB >>sliders. > >Well, that I'm not sure of. Jon, I do not mean restore the old system picker, but to add HSV fields to your new picker, right above the alpha spinbutton. _________________________________________________________________ Help STOP SPAM with the new MSN 8 and get 2 months FREE* http://join.msn.com/?page=dept/bcomm&pgmarket=en-ca&RU=http%3a%2f%2fjoin.ms…

2 1

RE: [Inkscape-devel] Some fairly random feature reqs + bug reports.
by bulia byak 13 Feb '04

13 Feb '04

Thanks for the feedback! > First, if I've got a new session of inkscape, and i >do file>open, why does it open it into a new window >rather than the blank one I've got in front of me? >every time it does it it strikes me as odd. Yes, it's in the plans. Maybe next version. > Second up, have you considered making use of the >boolean ops to add paintbrush/eraser tools ala flash? >the functionality is all there just spread across >several tools currently. The way I've been simulating >it currently is to use the calligraphy tool to 'paint' >an area, then whenever you paint additional >areas immediately switch to select and ctrl++ it to >the last section. Not sure about usefulness of this. The vector approach is to keep things separate and at a high level of abstraction for as long as possible, and only merge/flatten/convert to paths as the last stage if possible. So if I need this functionality, I would actually prefer to have shapes separate to I can move and adjust them, and do a union or difference on them if necessary (but I think in many cases, just grouping them together would be convenient enough). On the other hand, what you propose may be simple enough to implement, but we need to do a paint tool with user-definable strokes first, otherwise the booleans are of little use IMHO. > Another feature that I'd love to see would be >variable offsets, ie a dynamic offset with >node weighting allowing the offset to be modified at >each node so I could generate a line of varying >thickness (like the calligraphy tool) but still >editable by the basic centreline path. This would be a >godsend for doing comic style drawings where the line >weight is important. I know that ones a big task, just >putting the idea out there. Yes, useful, but perhaps more uesful in the short term and easier to implement is a variable-width stroke with some predefined shapes. I think our developers had some prototypes of this. > With regards to the 0.37 release, the builds i >downloaded from Bobs site (which is really useful, >both for builds and build instructions, cheers Bob) >seemed to have problems with the keyboard shortcuts >being messed up, the dialog ones were all muddled, >(for instance Ctrl+Shift+M was opening fill/stroke) >just checked the latest one there and its still the >case.(zip file Inkscape0402120319.zip) Very strange. Anyone else experiencing this? >What happened in the Fill/stroke dialog? the entry >boxs for RGB values on the wheel tab are gone That's temporary, we plan to restore them, although not RGB - the wheel is HSV and it will have HSV fields, if you need RGB switch to the RGB sliders. >as is >the dropper icon, both of which were really useful. Dropper on the dialog is redundant, you can use the dropper tool >(its a pain having to switch tool to do a dropper >action, What's the problem with the dropper? We discussed moving the dropper to the dialog and removing the tool, but then decided that the dropper will be very useful without the dialog, when we have a current color indicator and a palette. >btw, anyone got any suggestions as to the best place >to get started with getting familiar with the code? Read developer's howtos on Wiki - there's quite a lot of info there, although it's unorganized. Then look into the code and try to find what seems easiest - I don't know what it may be for you :) I started with adding new keyboard shortcuts, that was easy and fun, and gradually learned many parts of the codebase from there. _________________________________________________________________ Protect your PC - get McAfee.com VirusScan Online http://clinic.mcafee.com/clinic/ibuy/campaign.asp?cid=3963

3 2