index of table

[qgenesist]

I am on Lua 5.1 windows 64 bit.

I am using only this.
local serpent = require('serpent')
savF(_fp, (serpent.block(_famOBJ)))

savF simply writes the data from serpent.block to the file pointed to by _fp, and _famOBJ is my table that I want to save.
I use plugins in a program that is genealogy. The tables I will serialize are only 3 types.
'flat' tables.
tables within tables ... tables
perhaps function tables

the issue I am facing and either do not understand the options, or they dont exist is:

when I select a few random samples to send to your file in an indexed table it works fine I chose a run of 5 tables, indexed by 3, 5, 23, 30, 63 :

{
[3] = {
CHIL = {
{
BIRTH = "1980",
DEATH = " ",
ERA = "(1980 - )",
famc = {
{
fby = "",
fID = 3,
fpa = ""
} --[[table: 07CC8AC8]],
{
fby = "HUSB",
fID = 2314,
fpa = "ADOP"
} --[[table: 07CCB020]]
} --[[table: 07CD8018]],

the indexes to those individual tables [3] = and so on, are the family number which I will need for retrieval, brilliant, this is working and precisely the serialization I need.

Then, I choose the first 500 entries in my table, which do start at [1], and are sequential for quite a while, but are not actually sequential in the whole. The result.

{
{
CHIL = {
{
BIRTH = "1964",
DEATH = " ",
ERA = "(1964 - )",
famc = {
{
fby = "",
fID = 1,
fpa = ""
} --[[table: 07CD5FC0]]
} --[[table: 07CD0700]],
fams = {
{
DDAT = "",
fID = 56,
MDAT = "9 Jun 1984"
} --[[table: 07CD7988]]
} --[[table: 07CD6100]],

the indexing is gone, with exactly the same call to serpent.
I need to force that indexing, and as you may remember me from Zerobrane, I lack the requisite knowledge to do so.

HELP!!!!
Thank you Paul.

Ron Melby

[qgenesist]

I have this messed up, I need the ouput formatted like dump but the [3] = , [5]= of line.

[qgenesist]

It appears that if there is a table within the table first thing, that starts at index [1] then those interior table indexes are not included.

[qgenesist]

here is an abbreviated file of my problem.
I have saved a shortened copy of a working save. all well and fine, but now I have changed the index to 1...n instead of random numbers not starting at 1. I need this back with the keys of 1...n intact. serpent saves it without the 1...n keys at present.

_famOBJ.txt
i can give more information if needed.

[pkulchenko]

@qgenesist, I'm confused a bit. It looks like the file you are showing already has the indexes. If you are saying that you're like to have indexes 1,2,3 instead of not having indexes at all, then it shouldn't make any difference to the table that Lua produces (so the indexes are not included if they are sequential and start with 1). Why would you need them?

[qgenesist]

that is the file I have and want back, normally it would have no index, when I run it thru your code, I put the index there to show what I want.

so, since the table layout is not as I made it in the first place, (lua organizes the interior of the table as it pleases) there is no reason to believe that the table will be written out and retrieved and built in key order. I will not bore you with full code...I will shorthand it.

get ORD from record in fanc == -14
find the two fID in those records (index numbers now missing)
husb =tbl[fID];husb (let fID be 127)
wife = tbl[fID].wife (let fID be 256)

for k, v in pairs(husb.fams) do
if wife[fID] = husb.fams[fID] then
marr={husb[fID], wife{fID]
end
end
if next(marr) then

so.... how is it guarenteed in my 7500 entries that contain(some of them) 2000 table entries more (that I am calling records) that they will A, remain indexed as original, in original order, without the index andB) I can retrieve them numerically something like this above?

in the end of the matter, somewhere in my ordinal list though, there are missing numbers, but it is way out there past 500 or more table entries that is 1..500, 501.507, 512, 513....1795... it does not seem to catch that when I run the entire file (not a small undertaking) as I said, each of those empty arrays/tables in the attachment are beyond plumb full.