small comment utility.jl, some new functions to hilp with kanurangi data in Parse.jl

quietlight

Jan 13, 2025, 9:51 PM

7X4LD2DLYPE2OYEBYTMRKLBHDUIBOSJBXSTXIWX6QFAMC4VGBLKAC

Dependencies

[2] 6XJX6ED3 added parse
[3] BOPNWZL4 refactored Skraak.jl into sub files, tidy now
[4] L7G33K4C fixed a comment problem in Clips, and upgraded, now it compiles on Mac again
[5] 2UBDFCJH new files tracked
[6] QA2TJZRA ripped ML out into the SkraakML repo, compiles quicker now
[7] NV7FXZ5Q first commit
[*] PHGT4YMB extended Skraak.dddict, and work on all the functions that call it or make it

Change contents

file deletion: Parse.jl (----------)

[3.6598]→[2.6049:6081](∅→∅),[2.6081]→[2.3042:3042](∅→∅)

#=
a = ["201012_123456.wav", "201014_123456.WAV", "201217_123456.wav", "211122_123456.WAV"]

b = ["121020_123456.WAV", "141020_123456.wav", "171220_123456.WAV", "221121_123456.wav"]

c = ["20230609_103000.WAV", "20241109_201504.wav"]

d = [
"120119_003002.wav",
"180120_231502.wav",
"170122_010005.wav",
"010419_234502.WAV",
"310320_231502.wav",
"220824_231502.WAV",
"240123_231502.wav",
]
e = ["XYZ123_7689_20230609_103000.WAV", "string 20241109_201504.wav"]
f = [
"abcdefg__1234_180120_231502.wav",
"string 120119_003002.wav",
"ABCD EFG___170122_010005.wav",
"BHD_1234 010419_234502.WAV",
"cill xyz 310320_231502.wav",
"220824_231502.WAV",
"240123_231502.wav",
]
=#

using Dates
#Assumes it's getting a folder of 1 recording session, not random files.
#For 6 digit dates the year digits need to have less variation than the day digits.
#Does not handle miss shapen dates filenamesBad = ["20230609_1030qa.WAV", "20241109_20wp04.wav"]
function date_time_of_fname(filenames::Vector{String})
pattern = r"(\d{6}|\d{8})_\d{6}"
f2 = [match(pattern, str).match for str in filenames]
#date_time format must be the same for whole vector
g = map(x -> length(x), f2) |> x -> unique(x)
@assert length(g) == 1 "Different date formats in vector" #all same length
@assert length(f2[1]) == 15 || length(f2[1]) == 13 "Wrong length to be a date_time"
raw_dt = map(x -> (split(x, "_")[1], split(x, "_")[2]), f2)
d = parse_date_strings(first.(raw_dt))
t = parse_time_strings(last.(raw_dt))
dt = map((x, y) -> (x..., y...), d, t)
return map(x -> DateTime(x...), dt)
end

function parse_time_strings(times)
h = map(x -> tryparse(Int64, x[1:2]), times)
m = map(x -> tryparse(Int64, x[3:4]), times)
s = map(x -> tryparse(Int64, x[5:6]), times)
return zip(h, m, s) |> collect
end

function parse_date_strings(dates)
if length(dates[1]) == 8
y = map(x -> tryparse(Int64, x[1:4]), dates)
m = map(x -> tryparse(Int64, x[5:6]), dates)
d = map(x -> tryparse(Int64, x[7:8]), dates)
elseif length(dates[1]) == 6
y, m, d = parse_short_date_strings(dates)
else
error("Date is not 8 or 6 digits long")
end
return zip(y, m, d) |> collect
end

# 6 digit dates
function parse_short_date_strings(dates)
@assert length(dates) > 1 "Not enough files to work out YYMMDD v DDMMYY"
x1 = map(x -> tryparse(Int64, x[1:2]), dates)
m = map(x -> tryparse(Int64, x[3:4]), dates)
x2 = map(x -> tryparse(Int64, x[5:6]), dates)
if length(dates) > 1
length(unique(x2)) >= length(unique(x1)) ? (y, d) = ((x1 .+ 2000), x2) :
(y, d) = ((x2 .+ 2000), x1) #assumes year was in 2000's, if 1 file or variance in y and y is equal, it assumes DDMMYY
else
@info "Not enough files to work out YYMMDD v DDMMYY, assuming DDMMYY" #check this is sensible I think DDMMYY is most common
y, d = (x2 .+ 2000), x1
end
return y, m, d
end
edit in src/Utility.jl at line 117

[3.327184]

[9.341736]

#dr = Dates.Date(2024,10,1):Day(1):Dates.Date(2024,12,31);
replacement in src/Utility.jl at line 137

[3.327718]→[3.327718:327823](∅→∅)

"https://api.sunrise-sunset.org/json?lat=-45.50608&lng=167.47822&date=$day&formatted=0",

[3.327718]

[3.327823]

#"https://api.sunrise-sunset.org/json?lat=-45.50608&lng=167.47822&date=$day&formatted=0", #C05
"https://api.sunrise-sunset.org/json?lat=-40.84856&lng=172.80814&date=$day&formatted=0", #Takaka
edit in src/Parse.jl at line 81

[2.3003]

[2.3003]

end

function parse_start(filenames::Vector{String})
return map(x -> (split(x, "-")[end-1] |> x -> parse(Int64, x)), filenames)
edit in src/Parse.jl at line 86

[2.3007]

function parse_end(filenames::Vector{String})
return map(
x -> (split(x, "-")[end] |> x -> chop(x, tail = 4) |> x -> parse(Int64, x)),
filenames,
)
end

function call_start_time(datetimes::Vector{DateTime}, start::Vector{Int64})
return 0
end