Post by: The Librarian on August 10, 2019, 08:49:24 pm
First things first though - is everyone happy with the filled circle/ellipse code up there?
Post by: bplus on August 10, 2019, 09:20:08 pm
https://www.qb64.org/forum/index.php?topic=1511.0
First, I have been very happy with filled circle and ellipse code actually has come in handy for me, thanks!
Post by: SMcNeill on August 10, 2019, 09:29:12 pm
MEMSort: http://qb64.freeforums.net/thread/28/mem-sort
ExtendedTimer: http://qb64.freeforums.net/thread/20/extended-timer-uet
SaveImage: (needs updated on forums to the latest version and then linked)
POSIX File and Dir Listing to Arrays: http://qb64.freeforums.net/thread/63/file-directory-listing-arrays
Spell Checker: https://www.qb64.org/forum/index.php?topic=1587.0
MemToHex: https://www.qb64.org/forum/index.php?topic=1553.0
All of the above are handy little routines which I keep in my toolbox and have reused in a ton of different ways.
Post by: STxAxTIC on August 11, 2019, 04:39:32 am
I happen to know the Librarian connects through a mobile hotspot so updates may a bit slow.
The old "saveimage" routine is a gem Steevo! Resurrect that for us, I bet the newbies will be thrilled!
Post by: STxAxTIC on August 11, 2019, 05:53:25 am
So MEM is different than all QB64 commands, amirite? It's strange from a QB45 point of view, takes a minute to master from a QB64 point of view, and I've noticed that almost half of Steve's tools involve it.
So the question is, should we cram MEM-related stuff into its own category? Something all-inclusive to MEM, including tutorials. It may be sortof a turnoff to see all the MEM stuff scattered into everything else. --- OR, is this 100% off base? ... because Toolbox entries should be black boxes that shouldnt be looked into... hmmm... opinions?
Post by: Petr on August 11, 2019, 06:05:13 am
Please look to my previous answer https://www.qb64.org/forum/index.php?topic=1566.msg107875#msg107875
Post by: pforpond on August 11, 2019, 06:09:15 am
As the topic says, it's time we expand our Samples Gallery and Toolbox. Please reply to this post with links to existing threads in Programs with code that qualifies.
First things first though - is everyone happy with the filled circle/ellipse code up there?
I'm curious as to what 'code that qualifies' means :) Is there a criteria to me met?
Post by: STxAxTIC on August 11, 2019, 06:16:02 am
Quote
I'm curious as to what 'code that qualifies' means :) Is there a criteria to me met?
You got it m8.
https://www.qb64.org/forum/index.php?topic=178.0 (https://www.qb64.org/forum/index.php?topic=178.0)
https://www.qb64.org/forum/index.php?topic=1062.0 (https://www.qb64.org/forum/index.php?topic=1062.0)
Post by: pforpond on August 11, 2019, 06:33:08 am
You got it m8.
https://www.qb64.org/forum/index.php?topic=178.0 (https://www.qb64.org/forum/index.php?topic=178.0)
https://www.qb64.org/forum/index.php?topic=1062.0 (https://www.qb64.org/forum/index.php?topic=1062.0)
Thank you. Looks like I probably don't comment my code enough to qualify. I'd submit this (https://www.qb64.org/forum/index.php?topic=34.msg167#msg167) but I didn't post the source code in the thread and it was all spaghetti anyways :)
Post by: SMcNeill on August 11, 2019, 08:07:49 am
Not sure where to begin this discussion, but may as well be here.
So MEM is different than all QB64 commands, amirite? It's strange from a QB45 point of view, takes a minute to master from a QB64 point of view, and I've noticed that almost half of Steve's tools involve it.
So the question is, should we cram MEM-related stuff into its own category? Something all-inclusive to MEM, including tutorials. It may be sortof a turnoff to see all the MEM stuff scattered into everything else. --- OR, is this 100% off base? ... because Toolbox entries should be black boxes that shouldnt be looked into... hmmm... opinions?
_MEM really isn’t that strange; it’s QB64’s updated take on PEEK and POKE. Since “toolboxes” tend to contain those routines which you use over and over again, they tend to be as efficient as possible, so they’ll plug into any program — and you can’t beat the speed and performance of direct memory access I Which MEM gives. I think a lot of toolbox code will make use of MEM, when it can. Probably no reason to separate them into their own subfolder. ;)
Post by: bplus on August 11, 2019, 09:04:19 am
Not sure where to begin this discussion, but may as well be here.
So MEM is different than all QB64 commands, amirite? It's strange from a QB45 point of view, takes a minute to master from a QB64 point of view, and I've noticed that almost half of Steve's tools involve it.
So the question is, should we cram MEM-related stuff into its own category? Something all-inclusive to MEM, including tutorials. It may be sortof a turnoff to see all the MEM stuff scattered into everything else. --- OR, is this 100% off base? ... because Toolbox entries should be black boxes that shouldnt be looked into... hmmm... opinions?
Categories would be great!
Who wants to look for a saw in the sandpaper pile or a screwdriver in metric socket sets?
But can this be done on a forum? 0f course it can, if there is a Will to do it! ;-)
If categories are by function, MEM routines would span many categories for a speedier way to do a numbers crunching task, is my thought.
Post by: bplus on August 11, 2019, 09:18:36 am
We worked on this a bit here at the forum this is my version from my toolbox.
Code: QB64: [Select]
- 'notes: REDIM the array(0) to be loaded before calling Split '<<<<<<<<<<<<<<<<<<<<<<< IMPORTANT!!!!
- ' bplus modifications of Galleon fix of Bulrush Split reply #13
- ' http://xmaxw.[abandoned, outdated and now likely malicious qb64 dot net website - don’t go there]/forum/index.php?topic=1612.0
- ' this sub further developed and tested here: \test\Strings\Split test.bas
- copy = mystr 'make copy since we are messing with mystr when the delimiter is a space
- 'special case if delim is space, probably want to remove all excess space
- curpos = 1
- arrpos = 0
- arrpos = arrpos + 1
- curpos = dpos + LD
Function: it takes a string, say a whole file gulped in one go with a binary get, and parses the string into a string array using: the specified string to split, delimiter and array name to load.
My version handles splits delimited by spaces specially by removing extra spaces first. 1000 was added to name when Steve recommended only using REDIM _PRESERVE every 1000 times or so instead of at every split.
What else do you need? I have the code used to test the routine.
I am partial to handling the space delimiter specially because this was originally used for space only delimited parser for interpreters I played with but that can be removed and special filter applied in separate function.
Post by: STxAxTIC on August 11, 2019, 10:09:12 am
I think for now, until the collection proves unruly, that our existing tag system will work. Nay on categories. I'll "tell" the Librarian... :-)
Post by: STxAxTIC on August 11, 2019, 12:33:06 pm
Code: QB64: [Select]
- PRINT &B0
- PRINT &B1
- PRINT &B10
- PRINT &B11
- PRINT &B100
- 'does QB64 reverse &B numbers conver HEX$ for base 16, OCT$ for base 8, no base 2?
- INPUT "Enter integer to convert to binary "; test
- 'what is limit _integer64 can take
- i = 1000000000
- i = i + 1000000000
- PRINT i
- PRINT i
- 'maybe _BIN$
- n = nDec 'copy this because or nDec gets changed effecting main code
- p = p + 1
- n = n - d * bs ^ pow
- Dec2Base$ = b$
- i = integerBase10 'copy
- pow = 0
- b$ = "1" + b$
- i = i - 2 ^ pow
- b$ = "0" + b$
- pow = pow + 1
- BIN$ = "&B" + b$
Just want to make sure that:
(i) This is the proper link
(ii) Code in top post is the code you want to show
(iii) You know entering "0" for input causes the program to loop infinitely
Thanks!
------------------------------------------------------------------------------------------
EDIT:
Regarding Split1000: Yes, we need some dead-simple test code for that, 'specially given the huge warning at the top.
Post by: bplus on August 11, 2019, 01:21:04 pm
Here is just the BIN$ function with just the tester code for it.
(BTW that wasn't an infinite loop that was a test of _INTEGER64, I was testing a number of things with main code.)
Code: QB64: [Select]
- ' QB64 can reverse &H numbers with HEX$ and &O numbers with OCT$
- ' but what about &B numbers?
- ' Here is BIN$
- INPUT "(0 ends test) Enter a positive integer to convert to binary "; test
- 'BIN$ creates a string = the base 2 equivalent of the Base 10 positive integer
- 'and is prefixed with &B to work with VAL
- i = integerBase10 'copy
- pow = 0
- b$ = "1" + b$
- i = i - 2 ^ pow
- b$ = "0" + b$
- pow = pow + 1
- BIN$ = "&B" + b$
I will get going on split1000 next.
Post by: bplus on August 11, 2019, 02:51:49 pm
The "huge warning" is just reminder that the array to be loaded has to be a Dynamic String array base 0.
Here is bare essentials, save the code as Split.bas to test a bas file split into file lines:
Code: QB64: [Select]
- ' bplus modifications of Galleon fix of Bulrush Split reply #13 http://xmaxw.[abandoned, outdated and now likely malicious qb64 dot net website - don’t go there]/forum/index.php?topic=1612.0
- ' further modified with Steve McNeill's suggestion to REDIM _PRESERVE in clumps of 1000
- ' The Split SUB takes a given string and delimiter and breaks the string into a given Dynamic String array base 0.
- ' simple test modify string s$
- s$ = " ,1, 2, 3, 4 ,5 , 6 ,7, 8, " 'try start and finish 2 blanks
- INPUT "Next a test with this code file, press enter "; w$
- ' test with this bas file saved as "Split.bas"
- 'notes: the loadMeArray() needs to be dynamic string array with base 0 ie REDIM loadMeArray(0) AS STRING
- arrpos = arrpos + 1
- curpos = dpos + LD
Update: just ran a successful test reversing the order of the two tests loading the file into the array and then loading the smaller string into the array, without REDIM to erase.
Post by: STxAxTIC on August 12, 2019, 08:27:02 am
Post by: SMcNeill on August 12, 2019, 09:41:26 am
' QB64 can reverse &H numbers with HEX$ and &O numbers with OCT$
' but what about &B numbers?
' Here is BIN$
Negative numbers have a different representation depending on which variable type we're using.
-1 as a BYTE is &B11111111
-2 as a BYTE is &B11111110
-2 as a integer is &B1111111111111110
Which used to be hard to represent and get proper return values back for, but not anymore! I present to you an expanded BIN$ which you can use for both positive and negative values:
Code: QB64: [Select]
- n$ = bin$(12, 1)
- n$ = bin$(-2, 1) 'byte
- n$ = bin$(-2, 2) 'integer
- NEXT j, i1
Or, for those who like to keep BIN$ simple, the function could be easily broken down into 2 parts -- bin$ and nbin$ (negative bin$), as so:
Code: QB64: [Select]
I don't know if you guys would prefer either of these over the version Bplus posted, but I thought I'd offer them for you to look at.
Post by: bplus on August 12, 2019, 10:42:31 am
Ideally, if BIN$ could handle both positive and negative integers I'd prefer that but if I have to say how many bytes then definitely prefer just doing positive, so nBin$ with 2nd argument of bytes is great additional tool for those who might need that.
BTW I doubt HEX$ and OCT$ handle negatives either, so a tool for all 3 handling negatives?
Post by: bplus on August 12, 2019, 11:14:16 am
A couple of comments on the Split1000 entry to tool box, I hope you don't think this picky.
I did rename the sub simply Split, so the title for the tool might be just Split, though Split1000 has a catchy ring. :)
Your 2nd line in description:
Quote
Description:
Split receives a string (or entire text file) along with a delimiter argument (such as space, semicolon, tab, etc.) and prepares an array to store the string in fragments. The Nth member of the resulting array is the sub-string between the N-1 and Nth delimiter in the original string.
Did not ring quite right but got me thinking the number of delimiters and the number of strings loaded to the array.
It is true, if there are N delimiters then there are N+1 strings loaded into the zero based array including potentially empty strings to serve as place holders ie if the whole string is just N delimiters, the array will be loaded with N+1 empty strings.
What's not ringing right is that N usually denotes a count and counts usually start at 1. It is very important to stress Split uses a zero based array (and does not exactly prepare one).
It is also true that the last item in the array comes after the last delimiter in the string, counting wise that is N+1.
Thanks for reading carefully and considering this as alternate description:
Quote
'The Split SUB takes a given string and delimiter and loads a given Dynamic String array base 0 with parsed or empty strings s.t. if there are N delimiters there will be N+1 strings and the upper bound of the array will be N.
Post by: STxAxTIC on August 24, 2019, 04:33:47 pm
The Librarian has been slowly chipping away (well, the opposite really) at this archive... Next question is for Steve regarding saving images.
Would you say the save image stuff is ready to crystalize? Namely this post:
https://www.qb64.org/forum/index.php?topic=1605.0 (https://www.qb64.org/forum/index.php?topic=1605.0)
... but save GIF is not:
https://www.qb64.org/forum/index.php?topic=1618.0 (https://www.qb64.org/forum/index.php?topic=1618.0)
- OR -
Should we hold off until the zlib/inflate/deflate stuff is more thoroughly vetted?
Post by: SMcNeill on August 24, 2019, 04:45:49 pm
Hey fellas,
The Librarian has been slowly chipping away (well, the opposite really) at this archive... Next question is for Steve regarding saving images.
Would you say the save image stuff is ready to crystalize? Namely this post:
https://www.qb64.org/forum/index.php?topic=1605.0 (https://www.qb64.org/forum/index.php?topic=1605.0)
... but save GIF is not:
https://www.qb64.org/forum/index.php?topic=1618.0 (https://www.qb64.org/forum/index.php?topic=1618.0)
- OR -
Should we hold off until the zlib/inflate/deflate stuff is more thoroughly vetted?
SaveGIF works just fine and can be added anytime. Sooner or later, I’ll update it to allow animations, but probably not for quite some time. There’s also a SaveBMP and SavePNG library available as well, if someone just wants to save in one of those particular formats.
The SaveImage v2.1 Library works just fine and pulls all those libraries together into one place for us, as well as the JPG format. For Windows, it allows image saving in BMP, JPG, PNG, and GIF format. Linux doesn’t play nice with DECLARE LIBRARY with the zlib library, so PNG is excluded from it, but it still works with GIF, BMP, and GIF formats. (Which is part of the reason why I was working so hard to add the zlib library to my repo.)
Post by: SMcNeill on August 24, 2019, 05:20:06 pm
Code: QB64: [Select]
- 'Safely loads a font without destroying our current print location and making it revert to the top left corner.
- _FONT font#
- right = (right / _PRINTWIDTH(" ")) + 1 'convert the graphic X coordinate back to a monospace LOC column
- LOCATE down, right
Post by: STxAxTIC on August 24, 2019, 05:57:33 pm
Post by: SMcNeill on August 24, 2019, 06:23:25 pm
Right. I see a SUB up there, but a demo?
&!#@%!!
Sometimes, I think my clipboard is broken. I copied the routine from my personal set of "toolbox tools", and wrote a demo for it... Somehow, windows didn't bother to copy the whole demo, and I was left with just the copied sub to paste. Let's try this one more time:
Code: QB64: [Select]
- PRINT "First, I want to showcase the exisinging issue with the _FONT command."
- PRINT "Let's start out typing something, and then pause before loading a"
- PRINT "new font..."
- slowprint "NOW WHERE IS OUR PRINT CURSOR AT??"
- PRINT "Now, let's try this same type of thing, while using SafeLoadFont"
- PRINT "We'll start typing something, and then pause before safe loading"
- PRINT "a new font..."
- slowprint "Slowly watch what happens to our print cursor "
- SafeLoadFont font(1)
- slowprint "as we use SafeLoadFont to change "
- SafeLoadFont font(2)
- slowprint "fonts while happily printing to "
- SafeLoadFont 16
- slowprint "the screen, without a concern in the world! "
- PRINT "Now, isn't that something?!"
- _LIMIT 5
- 'Safely loads a font without destroying our current print location and making it revert to the top left corner.
- _FONT font#
- right = (right / _PRINTWIDTH(" ")) + 1 'convert the graphic X coordinate back to a monospace LOC column
- LOCATE down, right
(Do I need to load a copy of the fonts as well? They're fairly standard ones, which seem to ship with most every OS out there.)
Post by: bplus on August 27, 2019, 09:27:20 pm
I have been working with Ron on streamlining DATA and from the experience have improved our very useful Split SUB into an even more convenient tool to use by lifting the 0 base restriction from it's usage.
Here is an updated Demo that shows we are using base 1 instead of 0 and a new description for the Split SUB. I hope the Librarian will be pleased that we constantly try to improve our tools and sometimes succeed.
Code: QB64: [Select]
- 'new data setup structure to streamline code in setup
- 'globals seen inside SUBs and FUNCTIONs because SHARED
- 'VVVV REDIM means dynamic arrays, so can change in setup
- 'locals just seen in following main code section
- setup
- SUB setup '3rd method of data structure, this is meant to be edited over and over as add to a refine words and substitutes
- topIndex = 5 ' <<< make modifications to d() and then update this number, that's it!
- d(1) = "Months,January,February,March,April,May,June,July,August,September,October,November,December"
- d(2) = "Days,Monday,Tuesday,Wednesday,Thursday,Friday,Saturday,Sunday"
- d(3) = "Holidays,NewYears Day,MLK Day,Valentine's Day,Easter,Mother's Day,Memorial Day,Father's Day,Independence Day,Bplus Day,Labor Day,Halloween,Thanksgiving,Christmas"
- d(4) = "Test no further words/phrases in this line. (And the next test will do an empty string.)"
- d(5) = ""
- 'This SUB will take a given N delimited string, and delimiter$ and creates an array of N+1 strings using the LBOUND of the given dynamic array to load.
- 'notes: the loadMeArray() needs to be dynamic string array and will not change the LBOUND of the array it is given.
- arrpos = arrpos + 1
- curpos = dpos + LD
Description:
Quote
This SUB will take a given N delimited string, and delimiter$ and creates an array of N+1 strings using the LBOUND of the given dynamic array to load.
'
Post by: STxAxTIC on August 27, 2019, 09:44:42 pm
Post by: bplus on August 27, 2019, 10:12:45 pm
Another great tool I have used in the past (for evaluator for interpreter) is now new and improved and newly named! For instance it use to use a slow FOR structure and step through string 1 character at a time, now uses While loop and Instr to fly through, it is now also not restricted to just space delimited strings.
Here is the description:
Quote
'Description: Use Item$() Function to treat strings like arrays without having to use an array structure.
' This function does not throw a fit if you ask for an item number (index) it does not have, it just returns an empty string.
' In QB64, Functions can't return arrays through the function name, but they can return strings that the Item$() function can translate like an an array index. nItem numbers are the same as Counting numbers positive integers starting at 1. eg Item7$ = Item$(CommaDelimitedString$, 7, ",") 'get 7th Item in string
And here is the Demo, you might notice it is allot like the one for improved Split. That is because I was comparing the two because I couldn't decide which is better. Both are designed to be vast improvement over the old RESTORE, READ, DATA system.
Code: QB64: [Select]
- 'globals seen inside SUBs and FUNCTIONs because SHARED
- 'VVVV REDIM means dynamic arrays, so can change in setup
- 'locals just seen in following main code section
- setup
- j = 2
- j = j + 1
- SUB setup '3rd method of data structure, this is meant to be edited over and over as add to a refine words and substitutes
- topIndex = 6 ' <<< make modifications to d() and then update this number, that's it!
- d(1) = "Months,January,February,March,April,May,June,July,August,September,October,November,December"
- d(2) = "Days,Monday,Tuesday,Wednesday,Thursday,Friday,Saturday,Sunday"
- d(3) = "Holidays,NewYears Day,MLK Day,Valentine's Day,Easter,Mother's Day,Memorial Day,Father's Day,Independence Day,Bplus Day,Labor Day,Halloween,Thanksgiving,Christmas"
- d(4) = "Leap Years,2000,2004,2008,2012,2016,2020,2024,2028,2032,2036,2040,...,2096,2104,..."
- d(5) = "Test no further words/phrases in this line. (And the next test will do an empty string.)"
- d(6) = ""
- 'Description: Use Item$() Function to treat strings like arrays without having to use an array structure.
- ' This function does not throw a fit if you ask for an item number (index) it does not have, it just returns an empty string.
- ' In QB64, Functions can't return arrays through the function name, but they can return strings that the Item$() function can
- ' translate like an an array index. nItem numbers are the same as Counting numbers positive integers starting at 1.
- ' eg Item7$ = Item$(CommaDelimitedString$, 7, ",") 'get 7th Item in string
- c = c + 1
- c = c + 1
Update / Append:
So you may wonder when to use Split and when to use Item$, well it is more efficient to Split to an array if the string is huge, like a file, and it is vital reference for your application. On the other hand, Rons task: https://www.qb64.org/forum/index.php?topic=1655.0 is solved by getting info fast from potentially hundreds of smaller strings a sort of Just In Time processing that gives array like access to a string without actually having to create an array :)
Ha! we have 50 replies in that thread and I have yet to demo the perfect fit Item$ is for his task, stay tuned...
I also have the problem solved with a Split version too but... learning is step by step.
Post by: Qwerkey on August 28, 2019, 06:13:12 am
Librarian, please find attached. I believe that this would be suitable as a Sample Program. As well as showing off the good graphics which QB64 produces, this is an InForm-based game and advertises Fellippe's production.
Post:
https://www.qb64.org/forum/index.php?topic=632.msg5033#new (https://www.qb64.org/forum/index.php?topic=632.msg5033#new)