hackguide.doc revision 76726
1 2 A Hacker's Guide to NCURSES 3 4 Contents 5 6 * Abstract 7 * Objective of the Package 8 + Why System V Curses? 9 + How to Design Extensions 10 * Portability and Configuration 11 * Documentation Conventions 12 * How to Report Bugs 13 * A Tour of the Ncurses Library 14 + Library Overview 15 + The Engine Room 16 + Keyboard Input 17 + Mouse Events 18 + Output and Screen Updating 19 * The Forms and Menu Libraries 20 * A Tour of the Terminfo Compiler 21 + Translation of Non-use Capabilities 22 + Use Capability Resolution 23 + Source-Form Translation 24 * Other Utilities 25 * Style Tips for Developers 26 * Porting Hints 27 28 Abstract 29 30 This document is a hacker's tour of the ncurses library and utilities. 31 It discusses design philosophy, implementation methods, and the 32 conventions used for coding and documentation. It is recommended 33 reading for anyone who is interested in porting, extending or 34 improving the package. 35 36 Objective of the Package 37 38 The objective of the ncurses package is to provide a free software API 39 for character-cell terminals and terminal emulators with the following 40 characteristics: 41 * Source-compatible with historical curses implementations 42 (including the original BSD curses and System V curses. 43 * Conformant with the XSI Curses standard issued as part of XPG4 by 44 X/Open. 45 * High-quality -- stable and reliable code, wide portability, good 46 packaging, superior documentation. 47 * Featureful -- should eliminate as much of the drudgery of C 48 interface programming as possible, freeing programmers to think at 49 a higher level of design. 50 51 These objectives are in priority order. So, for example, source 52 compatibility with older version must trump featurefulness -- we 53 cannot add features if it means breaking the portion of the API 54 corresponding to historical curses versions. 55 56Why System V Curses? 57 58 We used System V curses as a model, reverse-engineering their API, in 59 order to fulfill the first two objectives. 60 61 System V curses implementations can support BSD curses programs with 62 just a recompilation, so by capturing the System V API we also capture 63 BSD's. 64 65 More importantly for the future, the XSI Curses standard issued by 66 X/Open is explicitly and closely modeled on System V. So conformance 67 with System V took us most of the way to base-level XSI conformance. 68 69How to Design Extensions 70 71 The third objective (standards conformance) requires that it be easy 72 to condition source code using ncurses so that the absence of 73 nonstandard extensions does not break the code. 74 75 Accordingly, we have a policy of associating with each nonstandard 76 extension a feature macro, so that ncurses client code can use this 77 macro to condition in or out the code that requires the ncurses 78 extension. 79 80 For example, there is a macro NCURSES_MOUSE_VERSION which XSI Curses 81 does not define, but which is defined in the ncurses library header. 82 You can use this to condition the calls to the mouse API calls. 83 84 Portability and Configuration 85 86 Code written for ncurses may assume an ANSI-standard C compiler and 87 POSIX-compatible OS interface. It may also assume the presence of a 88 System-V-compatible select(2) call. 89 90 We encourage (but do not require) developers to make the code friendly 91 to less-capable UNIX environments wherever possible. 92 93 We encourage developers to support OS-specific optimizations and 94 methods not available under POSIX/ANSI, provided only that: 95 * All such code is properly conditioned so the build process does 96 not attempt to compile it under a plain ANSI/POSIX environment. 97 * Adding such implementation methods does not introduce 98 incompatibilities in the ncurses API between platforms. 99 100 We use GNU autoconf(1) as a tool to deal with portability issues. The 101 right way to leverage an OS-specific feature is to modify the autoconf 102 specification files (configure.in and aclocal.m4) to set up a new 103 feature macro, which you then use to condition your code. 104 105 Documentation Conventions 106 107 There are three kinds of documentation associated with this package. 108 Each has a different preferred format: 109 * Package-internal files (README, INSTALL, TO-DO etc.) 110 * Manual pages. 111 * Everything else (i.e., narrative documentation). 112 113 Our conventions are simple: 114 1. Maintain package-internal files in plain text. The expected viewer 115 for them more(1) or an editor window; there's no point in 116 elaborate mark-up. 117 2. Mark up manual pages in the man macros. These have to be viewable 118 through traditional man(1) programs. 119 3. Write everything else in HTML. 120 121 When in doubt, HTMLize a master and use lynx(1) to generate plain 122 ASCII (as we do for the announcement document). 123 124 The reason for choosing HTML is that it's (a) well-adapted for on-line 125 browsing through viewers that are everywhere; (b) more easily readable 126 as plain text than most other mark-ups, if you don't have a viewer; 127 and (c) carries enough information that you can generate a 128 nice-looking printed version from it. Also, of course, it make 129 exporting things like the announcement document to WWW pretty trivial. 130 131 How to Report Bugs 132 133 The reporting address for bugs is bug-ncurses@gnu.org. This is a 134 majordomo list; to join, write to bug-ncurses-request@gnu.org with a 135 message containing the line: 136 subscribe <name>@<host.domain> 137 138 The ncurses code is maintained by a small group of volunteers. While 139 we try our best to fix bugs promptly, we simply don't have a lot of 140 hours to spend on elementary hand-holding. We rely on intelligent 141 cooperation from our users. If you think you have found a bug in 142 ncurses, there are some steps you can take before contacting us that 143 will help get the bug fixed quickly. 144 145 In order to use our bug-fixing time efficiently, we put people who 146 show us they've taken these steps at the head of our queue. This means 147 that if you don't, you'll probably end up at the tail end and have to 148 wait a while. 149 1. Develop a recipe to reproduce the bug. 150 Bugs we can reproduce are likely to be fixed very quickly, often 151 within days. The most effective single thing you can do to get a 152 quick fix is develop a way we can duplicate the bad behavior -- 153 ideally, by giving us source for a small, portable test program 154 that breaks the library. (Even better is a keystroke recipe using 155 one of the test programs provided with the distribution.) 156 2. Try to reproduce the bug on a different terminal type. 157 In our experience, most of the behaviors people report as library 158 bugs are actually due to subtle problems in terminal descriptions. 159 This is especially likely to be true if you're using a traditional 160 asynchronous terminal or PC-based terminal emulator, rather than 161 xterm or a UNIX console entry. 162 It's therefore extremely helpful if you can tell us whether or not 163 your problem reproduces on other terminal types. Usually you'll 164 have both a console type and xterm available; please tell us 165 whether or not your bug reproduces on both. 166 If you have xterm available, it is also good to collect xterm 167 reports for different window sizes. This is especially true if you 168 normally use an unusual xterm window size -- a surprising number 169 of the bugs we've seen are either triggered or masked by these. 170 3. Generate and examine a trace file for the broken behavior. 171 Recompile your program with the debugging versions of the 172 libraries. Insert a trace() call with the argument set to 173 TRACE_UPDATE. (See "Writing Programs with NCURSES" for details on 174 trace levels.) Reproduce your bug, then look at the trace file to 175 see what the library was actually doing. 176 Another frequent cause of apparent bugs is application coding 177 errors that cause the wrong things to be put on the virtual 178 screen. Looking at the virtual-screen dumps in the trace file will 179 tell you immediately if this is happening, and save you from the 180 possible embarrassment of being told that the bug is in your code 181 and is your problem rather than ours. 182 If the virtual-screen dumps look correct but the bug persists, 183 it's possible to crank up the trace level to give more and more 184 information about the library's update actions and the control 185 sequences it issues to perform them. The test directory of the 186 distribution contains a tool for digesting these logs to make them 187 less tedious to wade through. 188 Often you'll find terminfo problems at this stage by noticing that 189 the escape sequences put out for various capabilities are wrong. 190 If not, you're likely to learn enough to be able to characterize 191 any bug in the screen-update logic quite exactly. 192 4. Report details and symptoms, not just interpretations. 193 If you do the preceding two steps, it is very likely that you'll 194 discover the nature of the problem yourself and be able to send us 195 a fix. This will create happy feelings all around and earn you 196 good karma for the first time you run into a bug you really can't 197 characterize and fix yourself. 198 If you're still stuck, at least you'll know what to tell us. 199 Remember, we need details. If you guess about what is safe to 200 leave out, you are too likely to be wrong. 201 If your bug produces a bad update, include a trace file. Try to 202 make the trace at the least voluminous level that pins down the 203 bug. Logs that have been through tracemunch are OK, it doesn't 204 throw away any information (actually they're better than 205 un-munched ones because they're easier to read). 206 If your bug produces a core-dump, please include a symbolic stack 207 trace generated by gdb(1) or your local equivalent. 208 Tell us about every terminal on which you've reproduced the bug -- 209 and every terminal on which you can't. Ideally, sent us terminfo 210 sources for all of these (yours might differ from ours). 211 Include your ncurses version and your OS/machine type, of course! 212 You can find your ncurses version in the curses.h file. 213 214 If your problem smells like a logic error or in cursor movement or 215 scrolling or a bad capability, there are a couple of tiny test frames 216 for the library algorithms in the progs directory that may help you 217 isolate it. These are not part of the normal build, but do have their 218 own make productions. 219 220 The most important of these is mvcur, a test frame for the 221 cursor-movement optimization code. With this program, you can see 222 directly what control sequences will be emitted for any given cursor 223 movement or scroll/insert/delete operations. If you think you've got a 224 bad capability identified, you can disable it and test again. The 225 program is command-driven and has on-line help. 226 227 If you think the vertical-scroll optimization is broken, or just want 228 to understand how it works better, build hashmap and read the header 229 comments of hardscroll.c and hashmap.c; then try it out. You can also 230 test the hardware-scrolling optimization separately with hardscroll. 231 232 There's one other interactive tester, tctest, that exercises 233 translation between termcap and terminfo formats. If you have a 234 serious need to run this, you probably belong on our development team! 235 236 A Tour of the Ncurses Library 237 238Library Overview 239 240 Most of the library is superstructure -- fairly trivial convenience 241 interfaces to a small set of basic functions and data structures used 242 to manipulate the virtual screen (in particular, none of this code 243 does any I/O except through calls to more fundamental modules 244 described below). The files 245 246 lib_addch.c lib_bkgd.c lib_box.c lib_chgat.c lib_clear.c 247 lib_clearok.c lib_clrbot.c lib_clreol.c lib_colorset.c lib_data.c 248 lib_delch.c lib_delwin.c lib_echo.c lib_erase.c lib_gen.c 249 lib_getstr.c lib_hline.c lib_immedok.c lib_inchstr.c lib_insch.c 250 lib_insdel.c lib_insstr.c lib_instr.c lib_isendwin.c lib_keyname.c 251 lib_leaveok.c lib_move.c lib_mvwin.c lib_overlay.c lib_pad.c 252 lib_printw.c lib_redrawln.c lib_scanw.c lib_screen.c lib_scroll.c 253 lib_scrollok.c lib_scrreg.c lib_set_term.c lib_slk.c 254 lib_slkatr_set.c lib_slkatrof.c lib_slkatron.c lib_slkatrset.c 255 lib_slkattr.c lib_slkclear.c lib_slkcolor.c lib_slkinit.c 256 lib_slklab.c lib_slkrefr.c lib_slkset.c lib_slktouch.c lib_touch.c 257 lib_unctrl.c lib_vline.c lib_wattroff.c lib_wattron.c lib_window.c 258 259 are all in this category. They are very unlikely to need change, 260 barring bugs or some fundamental reorganization in the underlying data 261 structures. 262 263 These files are used only for debugging support: 264 265 lib_trace.c lib_traceatr.c lib_tracebits.c lib_tracechr.c 266 lib_tracedmp.c lib_tracemse.c trace_buf.c 267 268 It is rather unlikely you will ever need to change these, unless you 269 want to introduce a new debug trace level for some reasoon. 270 271 There is another group of files that do direct I/O via tputs(), 272 computations on the terminal capabilities, or queries to the OS 273 environment, but nevertheless have only fairly low complexity. These 274 include: 275 276 lib_acs.c lib_beep.c lib_color.c lib_endwin.c lib_initscr.c 277 lib_longname.c lib_newterm.c lib_options.c lib_termcap.c lib_ti.c 278 lib_tparm.c lib_tputs.c lib_vidattr.c read_entry.c. 279 280 They are likely to need revision only if ncurses is being ported to an 281 environment without an underlying terminfo capability representation. 282 283 These files have serious hooks into the tty driver and signal 284 facilities: 285 286 lib_kernel.c lib_baudrate.c lib_raw.c lib_tstp.c lib_twait.c 287 288 If you run into porting snafus moving the package to another UNIX, the 289 problem is likely to be in one of these files. The file lib_print.c 290 uses sleep(2) and also falls in this category. 291 292 Almost all of the real work is done in the files 293 294 hardscroll.c hashmap.c lib_addch.c lib_doupdate.c lib_getch.c 295 lib_mouse.c lib_mvcur.c lib_refresh.c lib_setup.c lib_vidattr.c 296 297 Most of the algorithmic complexity in the library lives in these 298 files. If there is a real bug in ncurses itself, it's probably here. 299 We'll tour some of these files in detail below (see The Engine Room). 300 301 Finally, there is a group of files that is actually most of the 302 terminfo compiler. The reason this code lives in the ncurses library 303 is to support fallback to /etc/termcap. These files include 304 305 alloc_entry.c captoinfo.c comp_captab.c comp_error.c comp_hash.c 306 comp_parse.c comp_scan.c parse_entry.c read_termcap.c write_entry.c 307 308 We'll discuss these in the compiler tour. 309 310The Engine Room 311 312 Keyboard Input 313 314 All ncurses input funnels through the function wgetch(), defined in 315 lib_getch.c. This function is tricky; it has to poll for keyboard and 316 mouse events and do a running match of incoming input against the set 317 of defined special keys. 318 319 The central data structure in this module is a FIFO queue, used to 320 match multiple-character input sequences against special-key 321 capabilities; also to implement pushback via ungetch(). 322 323 The wgetch() code distinguishes between function key sequences and the 324 same sequences typed manually by doing a timed wait after each input 325 character that could lead a function key sequence. If the entire 326 sequence takes less than 1 second, it is assumed to have been 327 generated by a function key press. 328 329 Hackers bruised by previous encounters with variant select(2) calls 330 may find the code in lib_twait.c interesting. It deals with the 331 problem that some BSD selects don't return a reliable time-left value. 332 The function timed_wait() effectively simulates a System V select. 333 334 Mouse Events 335 336 If the mouse interface is active, wgetch() polls for mouse events each 337 call, before it goes to the keyboard for input. It is up to 338 lib_mouse.c how the polling is accomplished; it may vary for different 339 devices. 340 341 Under xterm, however, mouse event notifications come in via the 342 keyboard input stream. They are recognized by having the kmous 343 capability as a prefix. This is kind of klugey, but trying to wire in 344 recognition of a mouse key prefix without going through the 345 function-key machinery would be just too painful, and this turns out 346 to imply having the prefix somewhere in the function-key capabilities 347 at terminal-type initialization. 348 349 This kluge only works because kmous isn't actually used by any 350 historic terminal type or curses implementation we know of. Best guess 351 is it's a relic of some forgotten experiment in-house at Bell Labs 352 that didn't leave any traces in the publicly-distributed System V 353 terminfo files. If System V or XPG4 ever gets serious about using it 354 again, this kluge may have to change. 355 356 Here are some more details about mouse event handling: 357 358 The lib_mouse()code is logically split into a lower level that accepts 359 event reports in a device-dependent format and an upper level that 360 parses mouse gestures and filters events. The mediating data structure 361 is a circular queue of event structures. 362 363 Functionally, the lower level's job is to pick up primitive events and 364 put them on the circular queue. This can happen in one of two ways: 365 either (a) _nc_mouse_event() detects a series of incoming mouse 366 reports and queues them, or (b) code in lib_getch.c detects the kmous 367 prefix in the keyboard input stream and calls _nc_mouse_inline to 368 queue up a series of adjacent mouse reports. 369 370 In either case, _nc_mouse_parse() should be called after the series is 371 accepted to parse the digested mouse reports (low-level events) into a 372 gesture (a high-level or composite event). 373 374 Output and Screen Updating 375 376 With the single exception of character echoes during a wgetnstr() call 377 (which simulates cooked-mode line editing in an ncurses window), the 378 library normally does all its output at refresh time. 379 380 The main job is to go from the current state of the screen (as 381 represented in the curscr window structure) to the desired new state 382 (as represented in the newscr window structure), while doing as little 383 I/O as possible. 384 385 The brains of this operation are the modules hashmap.c, hardscroll.c 386 and lib_doupdate.c; the latter two use lib_mvcur.c. Essentially, what 387 happens looks like this: 388 389 The hashmap.c module tries to detect vertical motion changes between 390 the real and virtual screens. This information is represented by the 391 oldindex members in the newscr structure. These are modified by 392 vertical-motion and clear operations, and both are re-initialized 393 after each update. To this change-journalling information, the hashmap 394 code adds deductions made using a modified Heckel algorithm on hash 395 values generated from the line contents. 396 397 The hardscroll.c module computes an optimum set of scroll, insertion, 398 and deletion operations to make the indices match. It calls 399 _nc_mvcur_scrolln() in lib_mvcur.c to do those motions. 400 401 Then lib_doupdate.c goes to work. Its job is to do line-by-line 402 transformations of curscr lines to newscr lines. Its main tool is the 403 routine mvcur() in lib_mvcur.c. This routine does cursor-movement 404 optimization, attempting to get from given screen location A to given 405 location B in the fewest output characters posible. 406 407 If you want to work on screen optimizations, you should use the fact 408 that (in the trace-enabled version of the library) enabling the 409 TRACE_TIMES trace level causes a report to be emitted after each 410 screen update giving the elapsed time and a count of characters 411 emitted during the update. You can use this to tell when an update 412 optimization improves efficiency. 413 414 In the trace-enabled version of the library, it is also possible to 415 disable and re-enable various optimizations at runtime by tweaking the 416 variable _nc_optimize_enable. See the file include/curses.h.in for 417 mask values, near the end. 418 419 The Forms and Menu Libraries 420 421 The forms and menu libraries should work reliably in any environment 422 you can port ncurses to. The only portability issue anywhere in them 423 is what flavor of regular expressions the built-in form field type 424 TYPE_REGEXP will recognize. 425 426 The configuration code prefers the POSIX regex facility, modeled on 427 System V's, but will settle for BSD regexps if the former isn't 428 available. 429 430 Historical note: the panels code was written primarily to assist in 431 porting u386mon 2.0 (comp.sources.misc v14i001-4) to systems lacking 432 panels support; u386mon 2.10 and beyond use it. This version has been 433 slightly cleaned up for ncurses. 434 435 A Tour of the Terminfo Compiler 436 437 The ncurses implementation of tic is rather complex internally; it has 438 to do a trying combination of missions. This starts with the fact 439 that, in addition to its normal duty of compiling terminfo sources 440 into loadable terminfo binaries, it has to be able to handle termcap 441 syntax and compile that too into terminfo entries. 442 443 The implementation therefore starts with a table-driven, dual-mode 444 lexical analyzer (in comp_scan.c). The lexer chooses its mode (termcap 445 or terminfo) based on the first `,' or `:' it finds in each entry. The 446 lexer does all the work of recognizing capability names and values; 447 the grammar above it is trivial, just "parse entries till you run out 448 of file". 449 450Translation of Non-use Capabilities 451 452 Translation of most things besides use capabilities is pretty 453 straightforward. The lexical analyzer's tokenizer hands each 454 capability name to a hash function, which drives a table lookup. The 455 table entry yields an index which is used to look up the token type in 456 another table, and controls interpretation of the value. 457 458 One possibly interesting aspect of the implementation is the way the 459 compiler tables are initialized. All the tables are generated by 460 various awk/sed/sh scripts from a master table include/Caps; these 461 scripts actually write C initializers which are linked to the 462 compiler. Furthermore, the hash table is generated in the same way, so 463 it doesn't have to be generated at compiler startup time (another 464 benefit of this organization is that the hash table can be in 465 shareable text space). 466 467 Thus, adding a new capability is usually pretty trivial, just a matter 468 of adding one line to the include/Caps file. We'll have more to say 469 about this in the section on Source-Form Translation. 470 471Use Capability Resolution 472 473 The background problem that makes tic tricky isn't the capability 474 translation itself, it's the resolution of use capabilities. Older 475 versions would not handle forward use references for this reason (that 476 is, a using terminal always had to follow its use target in the source 477 file). By doing this, they got away with a simple implementation 478 tactic; compile everything as it blows by, then resolve uses from 479 compiled entries. 480 481 This won't do for ncurses. The problem is that that the whole 482 compilation process has to be embeddable in the ncurses library so 483 that it can be called by the startup code to translate termcap entries 484 on the fly. The embedded version can't go promiscuously writing 485 everything it translates out to disk -- for one thing, it will 486 typically be running with non-root permissions. 487 488 So our tic is designed to parse an entire terminfo file into a 489 doubly-linked circular list of entry structures in-core, and then do 490 use resolution in-memory before writing everything out. This design 491 has other advantages: it makes forward and back use-references equally 492 easy (so we get the latter for free), and it makes checking for name 493 collisions before they're written out easy to do. 494 495 And this is exactly how the embedded version works. But the 496 stand-alone user-accessible version of tic partly reverts to the 497 historical strategy; it writes to disk (not keeping in core) any entry 498 with no use references. 499 500 This is strictly a core-economy kluge, implemented because the 501 terminfo master file is large enough that some core-poor systems swap 502 like crazy when you compile it all in memory...there have been reports 503 of this process taking three hours, rather than the twenty seconds or 504 less typical on the author's development box. 505 506 So. The executable tic passes the entry-parser a hook that immediately 507 writes out the referenced entry if it has no use capabilities. The 508 compiler main loop refrains from adding the entry to the in-core list 509 when this hook fires. If some other entry later needs to reference an 510 entry that got written immediately, that's OK; the resolution code 511 will fetch it off disk when it can't find it in core. 512 513 Name collisions will still be detected, just not as cleanly. The 514 write_entry() code complains before overwriting an entry that 515 postdates the time of tic's first call to write_entry(), Thus it will 516 complain about overwriting entries newly made during the tic run, but 517 not about overwriting ones that predate it. 518 519Source-Form Translation 520 521 Another use of tic is to do source translation between various termcap 522 and terminfo formats. There are more variants out there than you might 523 think; the ones we know about are described in the captoinfo(1) manual 524 page. 525 526 The translation output code (dump_entry() in ncurses/dump_entry.c) is 527 shared with the infocmp(1) utility. It takes the same internal 528 representation used to generate the binary form and dumps it to 529 standard output in a specified format. 530 531 The include/Caps file has a header comment describing ways you can 532 specify source translations for nonstandard capabilities just by 533 altering the master table. It's possible to set up capability aliasing 534 or tell the compiler to plain ignore a given capability without 535 writing any C code at all. 536 537 For circumstances where you need to do algorithmic translation, there 538 are functions in parse_entry.c called after the parse of each entry 539 that are specifically intended to encapsulate such translations. This, 540 for example, is where the AIX box1 capability get translated to an 541 acsc string. 542 543 Other Utilities 544 545 The infocmp utility is just a wrapper around the same entry-dumping 546 code used by tic for source translation. Perhaps the one interesting 547 aspect of the code is the use of a predicate function passed in to 548 dump_entry() to control which capabilities are dumped. This is 549 necessary in order to handle both the ordinary De-compilation case and 550 entry difference reporting. 551 552 The tput and clear utilities just do an entry load followed by a 553 tputs() of a selected capability. 554 555 Style Tips for Developers 556 557 See the TO-DO file in the top-level directory of the source 558 distribution for additions that would be particularly useful. 559 560 The prefix _nc_ should be used on library public functions that are 561 not part of the curses API in order to prevent pollution of the 562 application namespace. If you have to add to or modify the function 563 prototypes in curses.h.in, read ncurses/MKlib_gen.sh first so you can 564 avoid breaking XSI conformance. Please join the ncurses mailing list. 565 See the INSTALL file in the top level of the distribution for details 566 on the list. 567 568 Look for the string FIXME in source files to tag minor bugs and 569 potential problems that could use fixing. 570 571 Don't try to auto-detect OS features in the main body of the C code. 572 That's the job of the configuration system. 573 574 To hold down complexity, do make your code data-driven. Especially, if 575 you can drive logic from a table filtered out of include/Caps, do it. 576 If you find you need to augment the data in that file in order to 577 generate the proper table, that's still preferable to ad-hoc code -- 578 that's why the fifth field (flags) is there. 579 580 Have fun! 581 582 Porting Hints 583 584 The following notes are intended to be a first step towards DOS and 585 Macintosh ports of the ncurses libraries. 586 587 The following library modules are `pure curses'; they operate only on 588 the curses internal structures, do all output through other curses 589 calls (not including tputs() and putp()) and do not call any other 590 UNIX routines such as signal(2) or the stdio library. Thus, they 591 should not need to be modified for single-terminal ports. 592 593 lib_addch.c lib_addstr.c lib_bkgd.c lib_box.c lib_clear.c 594 lib_clrbot.c lib_clreol.c lib_delch.c lib_delwin.c lib_erase.c 595 lib_inchstr.c lib_insch.c lib_insdel.c lib_insstr.c lib_keyname.c 596 lib_move.c lib_mvwin.c lib_newwin.c lib_overlay.c lib_pad.c 597 lib_printw.c lib_refresh.c lib_scanw.c lib_scroll.c lib_scrreg.c 598 lib_set_term.c lib_touch.c lib_tparm.c lib_tputs.c lib_unctrl.c 599 lib_window.c panel.c 600 601 This module is pure curses, but calls outstr(): 602 603 lib_getstr.c 604 605 These modules are pure curses, except that they use tputs() and 606 putp(): 607 608 lib_beep.c lib_color.c lib_endwin.c lib_options.c lib_slk.c 609 lib_vidattr.c 610 611 This modules assist in POSIX emulation on non-POSIX systems: 612 613 sigaction.c 614 signal calls 615 616 The following source files will not be needed for a 617 single-terminal-type port. 618 619 alloc_entry.c captoinfo.c clear.c comp_captab.c comp_error.c 620 comp_hash.c comp_main.c comp_parse.c comp_scan.c dump_entry.c 621 infocmp.c parse_entry.c read_entry.c tput.c write_entry.c 622 623 The following modules will use open()/read()/write()/close()/lseek() 624 on files, but no other OS calls. 625 626 lib_screen.c 627 used to read/write screen dumps 628 629 lib_trace.c 630 used to write trace data to the logfile 631 632 Modules that would have to be modified for a port start here: 633 634 The following modules are `pure curses' but contain assumptions 635 inappropriate for a memory-mapped port. 636 637 lib_longname.c 638 assumes there may be multiple terminals 639 640 lib_acs.c 641 assumes acs_map as a double indirection 642 643 lib_mvcur.c 644 assumes cursor moves have variable cost 645 646 lib_termcap.c 647 assumes there may be multiple terminals 648 649 lib_ti.c 650 assumes there may be multiple terminals 651 652 The following modules use UNIX-specific calls: 653 654 lib_doupdate.c 655 input checking 656 657 lib_getch.c 658 read() 659 660 lib_initscr.c 661 getenv() 662 663 lib_newterm.c 664 lib_baudrate.c 665 lib_kernel.c 666 various tty-manipulation and system calls 667 668 lib_raw.c 669 various tty-manipulation calls 670 671 lib_setup.c 672 various tty-manipulation calls 673 674 lib_restart.c 675 various tty-manipulation calls 676 677 lib_tstp.c 678 signal-manipulation calls 679 680 lib_twait.c 681 gettimeofday(), select(). 682 _________________________________________________________________ 683 684 685 Eric S. Raymond <esr@snark.thyrsus.com> 686 687 (Note: This is not the bug address!) 688