Mercurial > hg > xemacs-beta
comparison src/README @ 0:376386a54a3c r19-14
Import from CVS: tag r19-14
| author | cvs |
|---|---|
| date | Mon, 13 Aug 2007 08:45:50 +0200 |
| parents | |
| children | e45d5e7c476e |
comparison
equal
deleted
inserted
replaced
| -1:000000000000 | 0:376386a54a3c |
|---|---|
| 1 This directory contains the source files for the C component of XEmacs. | |
| 2 Nothing in this directory is needed for using XEmacs once it is built | |
| 3 and installed, if the dumped Emacs (on Unix systems) or the XEmacs | |
| 4 executable and map files (on VMS systems) are copied elsewhere. | |
| 5 | |
| 6 See the files ../README and then ../INSTALL for installation instructions. | |
| 7 | |
| 8 Under Unix, the file `Makefile.in.in' is used as a template by the script | |
| 9 `../configure' to produce `Makefile.in'. The same script then uses `cpp' | |
| 10 to produce the machine-dependent `Makefile' from `Makefile.in'; | |
| 11 `Makefile' is the file which actually controls the compilation of | |
| 12 Emacs. Most of this should work transparently to the user; you should | |
| 13 only need to run `../configure', and then type `make'. | |
| 14 | |
| 15 See the file VMSBUILD in this directory for instructions on compiling, | |
| 16 linking and building Emacs on VMS. | |
| 17 | |
| 18 The files `*.com' and `temacs.opt' are used on VMS only. | |
| 19 The files `vlimit.h', `ioclt.h' and `param.h' are stubs to | |
| 20 allow compilation on VMS with the minimum amount of #ifdefs. | |
| 21 | |
| 22 `uaf.h' contains VMS uaf structure definitions. This is only needed if | |
| 23 you define READ_SYSUAF. This should only be done for single-user | |
| 24 systems where you are not overly concerned with security, since it | |
| 25 either requires that you install Emacs with SYSPRV or make SYSUAF.DAT | |
| 26 world readable. Otherwise, Emacs can determine information about the | |
| 27 current user, but no one else. | |
| 28 | |
| 29 `pwd.h' contains definitions for VMS to be able to correctly simulate | |
| 30 `getpwdnam' and `getpwduid'. | |
| 31 | |
| 32 General changes for XEmacs: | |
| 33 --------------------------- | |
| 34 1. Lisp objects. | |
| 35 | |
| 36 -- XFASTINT has been eliminated. Use of this expression as an lvalue | |
| 37 is incompatible with the union form of Lisp objects, and use as | |
| 38 an rvalue is likely to lead to errors and doesn't really save much | |
| 39 time. Expressions of the form `XFASTINT (obj) = num;' get replaced | |
| 40 by `obj = make_int (num);' or `XSETINT (obj, num);' and | |
| 41 expressions of the form `num = XFASTINT (obj);' get replaced by | |
| 42 `num = XINT (obj);'. Use Qzero in place of `make_int (0)'. | |
| 43 | |
| 44 -- Use of XTYPE gets replaced by the appropriate predicate. Using | |
| 45 XTYPE only works for the small number of types that are not stored | |
| 46 using the Lisp_Record type (int, cons, string, and vector). For | |
| 47 example, `(XTYPE (foo) == Lisp_Buffer)' gets replaced by | |
| 48 `(BUFFERP (foo))'. | |
| 49 | |
| 50 -- `XSET (obj, Lisp_Int, num)' gets replaced by `XSETINT (obj, num)', | |
| 51 for consistency. | |
| 52 | |
| 53 -- Some occurrences of XSET need to get replaced by XSETR -- | |
| 54 specifically, those where the type is not a primitive type | |
| 55 (primitive types are int, cons, string, and vector). | |
| 56 | |
| 57 -- References to `XSTRING (obj)->size' get replaced with | |
| 58 `string_length (obj)'. This is currently for cosmetic reasons | |
| 59 but there may be other reasons in the future. (This change is | |
| 60 currently incomplete in the source files.) | |
| 61 | |
| 62 | |
| 63 2. Storage classes: | |
| 64 | |
| 65 -- All occurrences of `const' should get replaced by CONST. This | |
| 66 is to work around a header conflict with X11R4. | |
| 67 | |
| 68 -- All occurrences of `register' should be replaced by `REGISTER'. | |
| 69 It interferes with backtraces so we disable it if DEBUG_XEMACS | |
| 70 is defined. | |
| 71 | |
| 72 | |
| 73 3. Errors, messages, I18N3 snarfing: | |
| 74 | |
| 75 -- Errors are continuable in XEmacs but are not in FSF Emacs. | |
| 76 Therefore, it's important that functions do something reasonable | |
| 77 if an error gets continued. If you want to signal a non- | |
| 78 continuable error, the call to Fsignal() gets put inside a | |
| 79 `while (1)' loop. To facilitate this, and also for proper | |
| 80 I18N3 message snarfing, most calls to Fsignal() have been | |
| 81 replaced by calls to signal_error(), signal_simple_error(), | |
| 82 etc. | |
| 83 | |
| 84 -- Constant strings occurring in source files need to get wrapped | |
| 85 in a call to GETTEXT (or if inside of a call to `build_string', | |
| 86 change that function to `build_translated_string') if they don't | |
| 87 occur in certain places where the I18N3 message snarfer will see | |
| 88 them. For a complete discussion of this, see the file | |
| 89 lib-src/make-msgfile.l. | |
| 90 | |
| 91 -- Calls to `fprintf (stderr, ...)' and `printf (...)' get replaced | |
| 92 with calls to `stderr_out' and `stdout_out'. This is for I18N3 | |
| 93 message snarfing. | |
| 94 | |
| 95 4. Initialization: | |
| 96 | |
| 97 -- FSF constructs like `obj = intern ("string"); staticpro (&obj);' | |
| 98 get replaced by `defsymbol (&obj);'. This is for code cleanness | |
| 99 and better purespace usage. | |
| 100 -- FSF constructs like | |
| 101 obj = intern ("error"); | |
| 102 Fput (obj, Qerror_message, "message"); | |
| 103 Fput (obj, Qerror_conditions, some list); | |
| 104 get replaced by calls to deferror(). See the definition of | |
| 105 deferror() for how the correct arguments to pass. This is for | |
| 106 code cleanness and I18N3 message snarfing. | |
| 107 -- Code in keys_of_foo() functions has been moved into Lisp. |
