This is an example from the file utts: The first record identifies the file type and hence the structure of the rest of the file. The file is rewritten after every session, so the values change appropriately. The first character identifies whether recognition or recall is currently active. The second character identifies the ladder rung with which to commence a session The third character indicates whether affixes or primitives are being studied (Macteach 2&3) or unused (Macteach 1&4). The fourth character indicates the number of correct repetitions required to escape the error box after an incorrect response. The fifth character encodes the time alloted for a response in test mode (up to about 9 seconds to the nearest 10th. The sixth character indicates the file type. The following character is a return (Macteach 2&3) or the beginning of a string giving the file name of the comment file (Macteach 1&4) The remaining description is file dependent The first 4 bytes of the first field contain the rung numbers for both recognition and recall for affix mode and primitive mode (Macteach 2&3) Only two are used for Macteach 1&4. The value also indicates, if the flashcard is in the error box, if it is ready for release. Macteach 1&4 have an additional byte indicating which test mode is active (Sentences are tested in segments with words being replaced by blanks. This byte indicates what must be blanked. Following these bytes is the Loglan sentence. The next record is the English translation without the initial number bytes. A number in this position is the ordinal of an explanatory comment from the comment file. Here is the commentary that explains the parsing of the answer: /* Do a string comparison, word by word ignoring case and taking into account all possible alternatives. On entry str2 contains the correct string, perhaps with numerous alternatives, and str1 is the test string entered by the user. Str2 may contain numerous punctuation marks with special meanings. Parenthesized words (Š) indicate something which need not be present but may be. The marks = or : indicate the left hand end of a set of alternatives, = includes a space (The man=bought/purchased a coat) and : does not (He is a grand:father/parent) The marks + and & are right-hand ends, + including a space, and & not. Such right-hand ends are considered a part of each alternative. Alternative expressions are separated by '/'. A backslash is used to mark a punctuation character which must be present in str1, otherwise punctuation is ignored. A space ends a word or pseudo-word group, an entity which is compared as a unit. Square brackets indicate comments in the Loglan, comments in the English are separated by '|'. The match works by creating a line containing the text, and a pointer line which indicates transfer points to resume testing after a match fails.*/ Consider the example sentence I+hope-you/wish-you-would/would-like-you-to eat that. A correct answer would contain ³I², then one of the alternatives ³hope you², ³wish you would², or ³would like you to², followed by ³eat that². 1122N3M1-Text 010012 1 010012 2 010012 3 010012 4 010012 5 010012 6 040212Hompi ti Drink this! 040312Eo titci ti Please eat this. 040312Katca ta eo Watch that, please. 040312Klipu eo ti Keep, please, this. 030312Ao hompi ta I-want-you-to drink that. 040312Ae titci ta I+hope-you/wish-you-would/would-like-you-to eat that. An example of the prims file is as follows: 1124N1 4400badjo baj bough/limb/branch of.. 4400badlo bad bundle/package/parcel/pack of/containing.. 4400bakso bao box/crate/carton of/containing.. 4400bakto bak bucket/pail of/containing.. The first record is as described above. Remaining records are divided into several fields. The primitive is included with the rung values in the first field. The second field contains the affix or affixes, with a hyphen if none are assigned. The next field is the English response with alternatives as above, and the final field gives a brief indication of the place structure