a9d4d35fc1
Better has_uniform_structure.
2017-10-22 12:29:43 +02:00
ef0a4c82ed
Rather more robust nr_elements_in_seq.
2017-10-22 12:23:19 +02:00
5e657c6504
More robust super_tostring.
2017-10-22 11:55:05 +02:00
79b036e51e
Fixed bugs relating to weird __tostring metamethods.
2017-10-21 13:24:33 +02:00
45d3958e17
Added new common file, where DISPLAY variable has been moved to.
2017-10-09 13:54:56 +02:00
f332d6bf95
Major restructure in how tables are formatted. Options max_depth and recursion has been deprecated.
2017-07-25 17:49:06 +02:00
33fb88fdd7
What is an is not an identifier is now automatically determined by is_identifier.
2017-07-20 18:58:06 +02:00
93beb4bd1e
Finally implemented column alignment
...
It's currently a bit limited in scope, but looks great when active.
2017-07-20 12:01:46 +02:00
d0d8e6476f
Bunch a' thoughts and documentation.
2017-07-17 21:33:11 +02:00
4a41111d1e
analyse_structure is not automatically called, if no info regarding a table could be found.
2017-06-24 18:36:01 +02:00
8d88f6195d
analyse_structure can now extend a previously used info table.
2017-06-24 17:59:55 +02:00
72f1ba5dfd
Fixed some issues where pretty would create new global variables.
2017-06-16 17:04:14 +02:00
b37a02c850
Now avoids analysing tables deeper than max_depth.
2017-06-15 19:02:54 +02:00
1bfd4abb96
Added is_uniform predicate to analyse_structure.lua.
2017-06-10 23:17:30 +02:00
4e7a1a9839
Improved error handling in pretty.
2017-06-09 16:20:39 +02:00
08e659fc4a
Created temporary stopgap for when __index throws errors.
2017-06-09 15:49:15 +02:00
4d3cf18797
Refactored analyse_structure, and improved the quality of gotten depths.
2017-06-06 00:24:24 +02:00
39dad0dadb
analyse_structure refactored to perform breath-first traversal. It also now keeps track of breath first depth of tables.
2017-06-06 00:02:34 +02:00
9050d53962
Removed unused functionallity, and associated tests.
2017-04-14 13:09:05 +02:00
87b12e15b2
Fixed issues with analýse_structure.lua
...
The most significant was a misclassification of tables with number
exclusive keys, with huge jumps between each one.
Example: { 1, [300] = 300 }
This was caused by logic faults in `get_table_info`, which has been fixed.
When solving above issue, a two new issues appeared:
1. Due to a logic fault, the count of number of elements in a table was wrong,
leading to issues with tables with boolean keys. This was fixed with simple
logic changes.
2. Some very small tables (1 element) would be classified as sets, and thus
rendered wrongly. This was fixed by adding a MINIMUM_NUMBER_OF_SET_ELEMENTS
constant.
2017-04-14 13:06:43 +02:00
155c877987
Alternative process for determining short tables
...
This one is based on the representative width of the table. Not only
does this produce better results, but it's also more futureproof.
2017-04-14 12:19:23 +02:00
5bee4a7378
Small holes in sequences will now be filled with nil, rather than triggering a reclassification to PURE_MAP.
2017-04-14 10:56:38 +02:00
ec3ba2cab9
Improved error messages, and fixed issue with importing from other modules
2017-04-05 12:21:43 +02:00
0a79b830ba
Moved some constants into their own file, more info is stored in l, and improved some marker stuff.
2017-04-03 11:24:51 +02:00
c75ee00f88
Added shortness info-gathering to analyze_structure
2017-01-19 14:23:34 +01:00
e2570ee780
Working on more extensive table info gathering.
2017-01-19 14:02:34 +01:00
