inform7/inform6

Inform 6

Inform 6 is a much earlier, more conventional programming language than Inform 7. It is still used by the Inform toolchain, so it needs to be part of this repository, but this is not its real home.

Compiler

The directory |inform6| contains the most recent stable release made by the Inform 6 project, which is managed by David Kinder. In an informal way, Andrew Plotkin is the lead developer.

The makefile for building this is inform6.mk, and places the executable in the directory inform6/Tangled, just as if Inform 6 were an Inweb project - though in fact it is not. It pre-dates everything else in the Inform toolchain, and is not a literate program. To build just I6, set the working directory to the outer inform6 directory, and then:

make -f inform6/inform6.mk
inform6/Tangled/inform6 -h

Library

Stand-alone users of Inform 6 need a "library" of standing code in order to get very much done: but the Inform 7 toolchain doesn't, and so no library is included in this repository.

See inform6lib, or if you have a minimalist bent, perhaps PunyInform.

Test suite

A full-scale suite of tests for the Inform 6 compiler has been created by Andrew Plotkin, and lives here: Inform 6 compiler tests.

This repository contains a conversion of that suite to use the Intest tool, in line with other utilities in the Inform toolchain. See the Intest manual, but:

../intest/Tangled/intest inform6 all

will run all (around 300) test cases, spreading the load across multiple cores for speed. For example:

../intest/Tangled/intest inform6 all
inform6 -> cases: [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71] [72] [73] [74] [75] [76] [77] [78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89] [90] [91] [92] [93] [94] [95] [96] [97] [98] [99] [100] [101] [102] [103] [104] [105] [106] [107] [108] [109] [110] [111] [112] [113] [114] [115] [116] [117] [118] [119] [120] [121] [122] [123] [124] [125] [126] [127] [128] [129] [130] [131] [132] [133] [134] [135] [136] [137] [138] [139] [140] [141] [142] [143] [144] [145] [146] [147] [148] [149] [150] [151] [152] [153] [154] [155] [156] [157] [158] [159] [160] [161] [162] [163] [164] [165] [166] [167] [168] [169] [170] [171] [172] [173] [174] [175] [176] [177] [178] [179] [180] [181] [182] [183] [184] [185] [186] [187] [188] [189] [190] [191] [192] [193] [194] [195] [196] [197] [198] [199] [200] [201] [202] [203] [204] [205] [206] [207] [208] [209] [210] [211] [212] [213] [214] [215] [216] [217] [218] [219] [220] [221] [222] [223] [224] [225] [226] [227] [228] [229] [230] [231] [232] [233] [234] [235] [236] [237] [238] [239] [240] [241] [242] [243] [244] [245] [246] [247] [248] [249] [250] 
inform6 -> problems: [251] [252] [253] [254] [255] [256] [257] [258] [259] [260] [261] [262] [263] [264] [265] [266] [267] [268] [269] [270] [271] [272] [273] [274] [275] [276] [277] [278] [279] [280] [281] [282] [283] [284] [285] [286] [287] [288] [289] [290] [291] [292] [293] [294] [295] [296] [297] [298] [299] [300] [301] [302] [303] [304] [305] [306] [307] 
All 307 tests succeeded (time taken 0:04, 16 simultaneous threads)

Individual cases, such as large_object_short_name_test, can be run like so:

../intest/Tangled/intest inform6 large_object_short_name_test

To see the actual shell commands invoked, run with -verbose:

../intest/Tangled/intest inform6 -verbose large_object_short_name_test

The test script is an Intest script called inform6.intest.

The tests themselves are divided in two: Test Cases contains those which are expected to compile successfully, that is, where inform6 is expected to return a successful exit code, and Test Errors, where it is expected to print errors and return an unsuccessful exit. (Note that mere warnings do not do that.)

Each individual test is its own source file: for example, "box_quote_test.inf". In Andrew Plotkin's original suite, the same source file is sometimes compiled in anything up to a dozen or so different ways, making for multiple tests from the same source file. But not here. Each individual test has its own name and its own file. For example, the original test "forwardproptest.inf" is split here into four: forwardproptest.inf, forwardproptest_ns-G.inf, forwardproptest_ns.inf and forwardproptest-G.inf. Any test case whose name ends "-G" is compiled to Glulx; all others are compiled to the Z-machine.

In addition, the test-setter can optionally specify two other things:

1. If the test is called NAME, the tester looks for _Settings/NAME.txt. If this file exists, it is expected to contain the command-line options to give to inform6. For example, it might read:

   -G -D $OMIT_UNUSED_ROUTINES=1

If no such file exists, then the settings are -G if the test name ends with -G, and nothing at all if it doesn't.

2. If the test is called NAME, the tester looks for _Scripts/NAME.txt. If this exists, it should be a list of typed commands, one per line, which are piped through into the story file interpreter. Many of the tests here make story files which print some text and then quit by themselves: a script is needed only if the story expects command input.

The process of testing goes like this:

1. Run inform6 on the source file.

2. Write the console output (stdout and stderr combined) into _Console_Actual/NAME.txt. This is expected to match _Console_Ideal/NAME.txt. If it doesn't, the test fails. If the ideal output does not exist, the test halts saying so. Running the test with the -bless option in Intest automatically trusts the actual console output and copies it over to become the ideal. (-curse removes this; -rebless updates it to a newer version.)

"Match" in this sense disregards (a) the inform6 version number as printed in the opening line of its output, (b) the inform6 compilation date, and (c) the duration in seconds of any abbreviation calculation output by the -u option. For example:

Inform 6.36 (24th January 2022)
Beginning calculation of optimal abbreviations...
Cross-reference table (70 entries) built...
Pass 1
Pass 1,    0/70 'tri' (7 occurrences)  (0.0000 seconds)	
...

is deemed to match:

Inform 6.39 (14th March 2027)
Beginning calculation of optimal abbreviations...
Cross-reference table (70 entries) built...
Pass 1
Pass 1,    0/70 'tri' (7 occurrences)  (0.0240 seconds)	
...

3. Write an md5 hash of the story file to _md5_Actual/NAME.txt. This is the same definition of "canonical hash" used by Andrew Plotkin's suite, which masks off header bytes of the story file to do with date and version. Note that the hashes produced here are identical to those produced by Andrew's Python script.

If the file _md5_Ideal/NAME.txt exists, the tester checks that the two match. If they do, the test stops with a pass. The story file must be correct - it is identical to a version known to be right. If they do not match, the test is a fail.

4. So we only continue if there was no ideal hash. The story file is run either through dumb-frotz or dumb-glulxe, command-line interpreters for our two virtual machines. A command script is piped in if one has been provided (see above). The transcript of output is stored in _Transcript_Actual/NAME.txt. This is expected to match _Transcript_Ideal/NAME.txt. If it doesn't, the test fails. If the ideal output does not exist, the test halts saying so. Running the test with the -bless option in Intest automatically trusts the actual transcript and copies it over to become the ideal. (-curse removes this; -rebless updates it to a newer version.)

The suite contains transcripts even for those tests where ideal hashes are provided: this is so that we can check that they are doing the right thing even after a change to the compiler invalidates those hashes.

Coverage and limitations

This test suite is a proper superset of the original, but please note:

1. large_opcode_text_test is not executed on dumb-frotz because the interpreter cannot cope with the quantity of text in question, and would halt with a fatal error. (But the test still checks the story file's md5.)

2. The two -k debugfile tests of Advent and the module compilation tests are currently omitted. These are, barring accidents, the only aspects of any test not faithfully reproduced from the original.