@mtxr Anyway If I want to contribute, where should I started? I have zero knowledge about VS Code extension. But I want to learn

Anyway @mtxr Will you set up patreon profile? I want to support you monthly.

Sorry @benyaminl. I've recently changed my job, so It was a hurry here.

I've just did. Here is the Patreon profile https://www.patreon.com/mteixeira

This autocomplete task it's a quite hard, I need to find a lib that parses the SQL, or we could write one. Do you know how to do that?

Sorry @benyaminl. I've recently changed my job, so It was a hurry here.

I've just did. Here is the Patreon profile https://www.patreon.com/mteixeira

This autocomplete task it's a quite hard, I need to find a lib that parses the SQL, or we could write one. Do you know how to do that?

Hm.. It use the language server or not? I have no idea to be honest for starting. Maybe some clue about it?

Yeah, we use language server. But taking a step back, I think would be better to create a library, them you could also publish it and help other projects.

The library should take an SQL query, parse it and tell us what is a keyword, table, column, schema and so on.

Furthermore, after this first step, it could get a bunch of queries, and map relationships between tables and columns.

What do you think?

I agree, just I think I need to read the VS code documentation first, for now I have no time for doing it. But I'll try.

Regards, Benyamin Limanto sent from my ASUS Fonepad 8

Pada tanggal Kam, 10 Jan 2019 21.34 Matheus Teixeira < notifications@github.com menulis:

Yeah, we use language server. But taking a step back, I think would be better to create a library, them you could also publish it and help other projects.

The library should take an SQL query, parse it and tell us what is a keyword, table, column, schema and so on.

Furthermore, after this first step, it could get a bunch of queries, and map relationships between tables and columns.

What do you think?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/mtxr/vscode-sqltools/issues/110#issuecomment-453116150, or mute the thread https://github.com/notifications/unsubscribe-auth/ACr9NLNT-XA5c2m-Ritu57rENEDGCgZQks5vB09ogaJpZM4YYVIh .

A quick hack: how about parsing the text for contained table names and limit column names to found ones? That should be more simple to implement than including a full SQL parser and gets 80% of the work done. Not correct all the time, but very helpful.

I also think about that，but when I see the code, I feel help less hahaha..

Regards, Benyamin Limanto sent from my ASUS Fonepad 8

Pada tanggal Sen, 21 Jan 2019 16.52 Korkman <notifications@github.com menulis:

A quick hack: how about parsing the text for contained table names and limit column names to found ones? That should be more simple to implement than including a full SQL parser and gets 80% of the work done. Not correct all the time, but very helpful.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/mtxr/vscode-sqltools/issues/110#issuecomment-456012331, or mute the thread https://github.com/notifications/unsubscribe-auth/ACr9NL_UAfbnLqypsz9E2yVVO0oI03hGks5vFY3SgaJpZM4YYVIh .

@mtxr is this the language server ? https://github.com/mtxr/vscode-sqltools/blob/master/packages/language-server/requests/completion/models.ts

Yes, all the code is here: https://github.com/mtxr/vscode-sqltools/tree/master/packages/language-server .

Those are the models for rendering suggestions, but you will probably need too.

Here is were the new completion logic should come: https://github.com/mtxr/vscode-sqltools/blob/master/packages/language-server/index.ts#L133

As you ca see, I just return all the items, that should be filtered bases on the query parsing.

@Korkman great suggestion! I was thinking about a more long term approach, but for sure your suggestion solves the problem right now. Can you and @benyaminl work together on this?

My next steps will be to provide oracle (PR #88) and sqlite (people are starving for that) #51 support.

I've started to try some parse and complete. It's referenced in the commit above. But it's really not that simple :(

I won't be able to complete this for v0.17. But is the next feature for sure

Sorry @benyaminl. I've recently changed my job, so It was a hurry here.

I've just did. Here is the Patreon profile https://www.patreon.com/mteixeira

This autocomplete task it's a quite hard, I need to find a lib that parses the SQL, or we could write one. Do you know how to do that?

the best way to is to use XML and use the dom engine..... sql is base on bloc statements so you load it in an xml file ..... the xml template will hold all logic operators

var xmlDocAdmistration = XMLDocument

function getAppXMLCredentials(){ xmlHttp = new XMLHttpRequest; xmlHttp.responseType = "document"; xmlHttp.open("GET",'./Adminstration.xml',true); // Force the response to be parsed as XML xmlHttp.overrideMimeType('text/xml') xmlHttp.responseType = "document"; xmlHttp.send(); xmlHttp.onreadystatechange = function() { if (xmlHttp.readyState == 4) {

   xmlDocAdmistration = xmlHttp.responseXML;//

you now have a load xml document in memory your sql template

the logic you retrieve the main node inner text as the sql statement it base on blocs so you load the words in the correct node element as text ....

select is a node name its text is Select it holds nested elements * in dom you simple add a node element ..... node elements hold attributes you can query each word as an element in a nested main node ..... its innertext is the sql

i want to build a table so you load the sql template it holds all syntax or a view this is the workhorse where you build reports from ..... where you combined tables.. read only reports.....

a select can have where clause so the template would need a nested node element ..... if its empty the inner-text off the main node show up blank ..... now JavaScript become simple to make its sql dom statements add insert remove and scan attribute get attributes under a switch list....

to build the xml template get very complex sql statement and start binding xml nodes around its words then test it with others that the temple will function for all sql brake downs.....

when you build dialog for input you point then to the xml template entry and all syntax is held in the template..... with a node reference ... to change the program around becomes simple you edit the template instructions .... for each attribute ..... your not playing with JavaScript it does what its suppose to do load the applicable node reference into the dialog.... so you avoid re-altering complex codes when a new sqlight adds new versions or drops features ...... so you keep your java-scripts as simple as possible to retrieve and send info in like data driver would..... switch statements avoid if statements where a switch statements are simple to add to...... a node attribute value select.... let xml act as the if operator..... a data island

when have a xml data island do you really need an intellicence drop down editor menus... no your simply treating sql statements as form and its island holds all syntax....... drag drop and set switches ..... so the end result would work like this ...... https://pxt.azureedge.net/blob/3a391831dd0f29344eb364be8a3c3cdeaa874d52/static/blog/lego/05-15-2018/editorgif.gif sql are just block statements where one can create a union queries or other types this way .... in the image you see how one can set up bock statements once the data island is built even to create tables with use of color codes ..... so you would only need widows for top for drag and drop and grid below to view its output..... and the db tree window..... extra window for raw sql and second one ready made JavaScript where it creates connect code like a web app i pull the snip code out and inject it into my web code....... like reports are hard to build yet with tool like this you get the job done in an hour not days.....

the user inter face eliminates mass dialogues needed for every type of operations..... your in control..... your simply drawing out your data off one dialog.... as programmer you eliminate mass code this way ......we now have a wizard build template environment to test complex calls to add and test new calls to submit to main sand box files.....

here how i recommend setting up an sql phrasing engine bata
vscode is built on V8 engine and will not be using sax or other type java libs but the v8 engine dom for speed the best way is to create a folder on git hub to share its run time files ... just html java script and XML to get on board we need the v8 engine to load up and we do this with electronjs app electron has the debugger we need and will load a local xml file into memory

Clone this repository

$ git clone https://github.com/atom/electron-quick-start

Go into the repository

$ cd electron-quick-start

Install dependencies and run the app

$ npm install && npm start

we need no other node modals we just want the debugger to build compatible vs-code to run

the app has on its menu bar a show and hide the debugger....... works grate.....it exposes the dom sin-tax calls to drill into the xml template

so we only need to share its primary files .... our html files can submit sql statements into the xml for testing.... and store our run time function via html forms

this way we are not playing games with what you have working as a some what productive tool

we dont need the sqlite running we use a db simulator file an array of tables.... it just to hard to setup compatible python lib in electron for bata testing and we want testers to play around with it.... so we run a JavaScript fake data base file..... a fake data tree like you have just use frames v8 will let do cross calls..... with frames...... and the frames server as a way to setup working logic transfer a sand box app

all css will work here also its the same engine as Vs-code a sandbox app folder the best setup would be to know vs-code v8 engine version with the correct electron version and with a db simulator file we then avoid python lib cmake files for version upgrades... for every tester ranting i cant debug a new version.... can you help me install python..lol..no

here is where the magic happens unlike json xml has parent child relations and has open sintax to work with.... so a child node can be created using html tags and the dom render engine ignores attributes so we can use class = "stylename" when we copy nodes from a xml file..... and place in a list tree html tag operation. or remove entry from the list..... so all we do is drag and drop.... will the parent node accept the child node...... yes or no we can pack in the syntax in this way xml for drop down menus or input fields for the attribute value. split string its entry and load.... we end up like this..... xml and html are combined in the xml file connected to its style nane..... https://pxt.azureedge.net/blob/3a391831dd0f29344eb364be8a3c3cdeaa874d52/static/blog/lego/05-15-2018/editorgif.gif the end results is a valid sql inner-text value of root node <--- one call no array building or trash code.... list tree html tags is the best way to template out an sql statement... you can see in the image its just a list tag being used to create complex operations..... so we need someone in styles for the list view .... no boot strap code just simple color style only for drag drop... we do not want overload the dom..... with css instructions to inject sql statement is simple you create word array from it then sent its word commands into rebuilding the template view.... a bit harder then building one.....you have to cross check its word entries

here is better way..... create call to open a catview in your extension https://code.visualstudio.com/api/extension-guides/webview we need the the dom phrasing engine and the v8 debugger your extension cant modify the dom.... yet we can do it this way......

Scripts and message passing# Webviews are just like iframes, which means that they can also run scripts. JavaScript is disabled in webviews by default, but it can easily re-enable by passing in the enableScripts: true option.

we can now send sql into the cat window..... from a data connection tree...... our bata window its just an isolated window to produces sql syntax from with loaded xml schema with 100% dom access

Thanks @julientype ! I'll take a look at this during the weekend.

Great job!

what we do is have the data store feed xml data structure to the cat window not json to keep simple and compatible..... 1 the store data xml feeder 2 the html list tree for multidimensional form insert inline operations 3 the xml data schema snips..... the return value is sql and as a data store you simply format to xml your not creating dom object your injecting tree data with correct xml tags for the load data schema dom object in the client and our list form tree will get table names and fields from to edit the selected template node inner-text value, the end result is one call to get the sql statement base on the user drag and drop operations

now we need to collect our xml call stack used for or JavaScript switch statement operations we dont have load new libs .....

here is a debugger image i simply create a brake point on the load xml file and start collecting calls i have a highlighted comment returning the node name popup.....

you can do the same thing from a cat window..... set the loaded file a local vs code folder so off line developers can still use it.....

now all you do is set up cat window with bata sql wizards on it and we run brake points to start building codes and schema data..... by opening the debug window on the cat window ..... this will take time to build but the primary file operation are simple to setup..... we can just send code snip to you to up the data modal files they are local here is better code or schema file ..... sql is not simple we can share progress calls by just adding live code comments in the scripts without running errors........ like before you publish a new set of files just open the debugger check for errors.....so we can toy around with it step by step to get its basic functions to move data around working..... like its list tree css its drag drop in and out its icons.... data support files for all versions and future versions .. all basic structures that we dont want to back track in code to suit a new feature.....our foundation code setup because this is bata we can support more then one xml schema author where we can test different versions a drop down list...... and authors can update without braking integrity rules missing syntax or syntax error.... like old to newer file testing what version can best build a complex sql statement.... we can keep users productive even with errors we let them manually edit the statement before they submit to the store..... till the syntax bugs are worked out.....so we have a list tree form and its below output statement where the user can still alter its errors or add missing elements..... bata

While it would be nice for this plugin to analyse sql and suggest object type to put i reckon it would take lot of work to make it decent. Why not just add some kind of prefix to tables and columns for users to decide what they want to auto complete . for example select 1,2,3 from table. [ctr+space] and autocomplete lists all tables etc . ASFAIK ! is not used in sql so you could just do select c![ctr+space] for column autocomplete and t! for table

the object is to draw out your sql off relational logic switching statements like one would work with svg vector images is xml box, circles, layers and so on... yet we only get a sql word array stored in database to redraw our form array snips..... building is the simple part..... like if you wanted to change a value and not have to rebuild one c![ctr+space] your trying to memorizes commands where one may work on a variety type databases and the user ends up checking and re checking correct intle menus commands where this wizard will drive every commands to you as an inline form entry ..... base on user selective operations .....
i need a new view what do you charge the client and how fast can you type it out..... the ax vs the chainsaw and your load dom xml can act as intel-a-menus also.... its just a loaded xml as a speed file array.... format format format in the xml you make it dance to your tune of operation and your in control of your css files not vscode where we sill get errors and wait....

here how simple this is.... dom object 1 html with its CSS pseudo-class <-JavaScript none needed ..its just appended dom call events https://developer.mozilla.org/en-US/docs/Web/CSS/Pseudo-classes dom object 2 xml data store for operations dom object 3 xml sql syntax forms javascript file A transponder code for the 3 dom object in operations it just takes a node and appends or removes or send values..... done..... domHTM needs to append node from xmlSQLDom node to its injected layout css responds domHtml need node from xmldomdata
and css responds domhtml form snips need to send values to its xmlsqldom template and returns the root node inner-text to domhtml to its sql view area value.....

is my java-script huge..... no.... its tiny..... its a transponder script... only...... drag drop remove and runs list form snips call... no arrays stored in java-scripts just a for each form snip code in xmlsgldom and add it to the domhtml as a menus selector structure pointer for drag drop and let css do its job.... and reprogram the xml array for html inputs without braking your JavaScript integration rules calls to xml naming and array structures..... or your back reediting your transponder scripts..... clever xml arrays make clever transponder scripts

one on layouts one on transponder scripts and one on xml data structures and your bata program is up and running....

mtxr is the data store programmer and needs help and loves to code and serve.. we need a teem..... for the debug and the construct of cat window he will set up.... he runs the site and cant do all........ so we brake it down css layout user define icons and so on good with colors and test script structure drag drop structures api calls to the dom xml sql builder expert in sql and known html and the master debugger to review and implement mtxr will feed code to the client cat window and improve the data store....

free help wanted and mtxr can trun it to a commercial app insert rewarding its team member coders like a share equality asset ...... where its members work at grantor good of the alll.. where 5 talented players are building a component project to leverage each others talent .... organized labor makes money not money or on sew string budget .... i pay this guy to program because i know not and his gone.... dirt code administration a spaghetti code component that holes no return value.. mtxr is the honest broker to create a component a team captain and it hockey players or it ends up a spaghetti nightmare on him self.... a team one is the honest broker to disbash ..... with it player to win the cup..... free help wanted.... i need the component to build for what i need....

when we have our 5 we close this help wanted rant and build....

best video ever to learn how to code and make money from.... https://www.youtube.com/watch?v=-sUXMzkh-jI i only need 5 people im only 1 mtxr is 2 who else can we count on ..help wanted do you get it options bangs once.... it cant bug you anymore if on your option no more then it express to its flower color to its direction.... ohh just an sql rant

my rants in the game is to provide my clients with tools you dont have....did i get paid to build them no.. can i yes

xml was created in 1998 v 1.... i was their working with developers do you see a new version 2 ..no... json is it predatory to shrink xml base from guys like me complaining to Microsoft its sintax is to big.... json is now made from 1998 we implanting.... ohh you use the same stuff today i know dom css from 1998 and its relay not change much on how to implement its v1 version its solid as sql from 1970 json is application boot structure tiny foot print xml is data island standard for applications development in elaborate context sql is data container retrieval language json will create more JavaScript where xml wont its a full dom array and json is just a boot script language ....

After failing the first attempt, I have good news: https://github.com/mtxr/sqltools-formatter/pull/1/files

Im adding the ability to identify table names to our formatter tool, its better because the formatter is non blocking and non validating. The other parsers I tried I need to handle when it validates a query like 'SELECT FROM table', all of them gave errors because of missing column names, and to workaround it was impossible.

Since query runner will throw error for invalid queries, I'm moving forward without validating it on-type.

This is a PR with just a starting code, needs to be worked a bit more, but seems promising!

After failing the first attempt, I have good news: https://github.com/mtxr/sqltools-formatter/pull/1/files Im adding the ability to identify table names to our formatter tool, its better because the formatter is non blocking and non validating. The other parsers I tried I need to handle when it validates a query like 'SELECT FROM table', all of them gave errors because of missing column names, and to workaround it was impossible. Since query runner will throw error for invalid queries, I'm moving forward without validating it on-type. This is a PR with just a starting code, needs to be worked a bit more, but seems promising!

Congratulation Mateo! Another milestones, huge one ! It will be the only sql tools that support multi database engine autocomplete on VS Code! Yey!

I've made some improvements on master. Just to allow easily pick tables for now, still a lot to do but it`s something.

The other branch will address most of column completions.

See v0.17.11

I'll be testing this today before publishing, but here is the link if you want to test

https://www.patreon.com/posts/v0-17-11-preview-25895548

Feedbacks are always welcome.

I'll be testing this today before publishing, but here is the link if you want to test https://www.patreon.com/posts/v0-17-11-preview-25895548 Feedbacks are always welcome.

It's a huge task! Take your time!

Hello @mtxr Did you have merged this feature into master ? I'm under 0.19.7 and I don't have this autocomplete filter.

Apologies, did not realise this was here (coming from #371)!!! Clearly ran a rubbish search 😆

Is there any ETA for this to land?

Not yet. I have to main issues for now. Auto complete and enhance the explorer to reduce memory consumption for large dbs.

I believe I did the hardest part so far from the auto complete, the tokenizer, now I need to map the tokens with the data that we already have. Problem is that I running out of time everyday :(

On Wed, Sep 4, 2019 at 8:42 AM willstocks notifications@github.com wrote:

Apologies, did not realise this was here (coming from #371 https://github.com/mtxr/vscode-sqltools/issues/371)!!! Clearly ran a rubbish search 😆

Is there any ETA for this to land?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/mtxr/vscode-sqltools/issues/110?email_source=notifications&email_token=AAFMX2PP6JHIEVNEWSUVQ5LQH6NINA5CNFSM4GDBKIQ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD53JCLY#issuecomment-527864111, or mute the thread https://github.com/notifications/unsubscribe-auth/AAFMX2MUXBLFHFWULMJI3A3QH6NINANCNFSM4GDBKIQQ .

Not yet. I have to main issues for now. Auto complete and enhance the explorer to reduce memory consumption for large dbs. I believe I did the hardest part so far from the auto complete, the tokenizer, now I need to map the tokens with the data that we already have. Problem is that I running out of time everyday :( … On Wed, Sep 4, 2019 at 8:42 AM willstocks @.***> wrote: Apologies, did not realise this was here (coming from #371 <#371>)!!! Clearly ran a rubbish search 😆 Is there any ETA for this to land? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#110?email_source=notifications&email_token=AAFMX2PP6JHIEVNEWSUVQ5LQH6NINA5CNFSM4GDBKIQ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD53JCLY#issuecomment-527864111>, or mute the thread https://github.com/notifications/unsubscribe-auth/AAFMX2MUXBLFHFWULMJI3A3QH6NINANCNFSM4GDBKIQQ .

It will be a year we celebrate this issue. Horray! keep on fire sir!

@benyaminl has funded $5.00 to this issue.

• Submit pull request via IssueHunt to receive this reward.
• Want to contribute? Chip in to this issue via IssueHunt.
• Checkout the IssueHunt Issue Explorer to see more funded issues.
• Need help from developers? Add your repository on IssueHunt to raise funds.

We got some progress on that. Please take a look at the new version

@mtxr oh it hasn't landed yet? Will you get the Price from IssueHunt? If yes I want to keep funding this issue.

Yes, it is. And no, I will not. Apparently that price would be received if someone else worked on that, you can request a refund

Yes, it is. And no, I will not. Apparently that price would be received if someone else worked on that, you can request a refund

Do you not interested in this progress :v *asking like a cat :v

@benyaminl has cancelled funding for this issue.(Cancelled amount: $5.00) See it on IssueHunt

what do you mean?

what do you mean?

nope just forget it. I only interested funding you monthly for this. Sorry if it seems rude.

hahaha no worries, I'm interested in funding for sure. But this issue as is had a lot of progress, so that's why I didn't understand your question

hahaha no worries, I'm interested in funding for sure. But this issue as is had a lot of progress, so that's why I didn't understand your question

I tried on my VS Code it still now showing anything when I do ctrl+space, is there anything that I need to do? I already connected to the database server and specify the DB name.

I'll dig into it today and get back to you

On Thu, Jun 11, 2020, 02:09 Benyamin Limanto notifications@github.com wrote:

hahaha no worries, I'm interested in funding for sure. But this issue as is had a lot of progress, so that's why I didn't understand your question

I tried on my VS Code it still now showing anything when I do ctrl+space, is there anything that I need to do? I already connected to the database server and specify the DB name.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/mtxr/vscode-sqltools/issues/110#issuecomment-642410330, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAFMX2MFXY5GDCMPZYAI62TRWBRHXANCNFSM4GDBKIQQ .

Thanks 🙏

Regards, Benyamin Limanto sent from my Sony XPERIA XZ Premium

On Thu, Jun 11, 2020, 19:56 Matheus Teixeira notifications@github.com wrote:

I'll dig into it today and get back to you

On Thu, Jun 11, 2020, 02:09 Benyamin Limanto notifications@github.com wrote:

hahaha no worries, I'm interested in funding for sure. But this issue as is had a lot of progress, so that's why I didn't understand your question

I tried on my VS Code it still now showing anything when I do ctrl+space, is there anything that I need to do? I already connected to the database server and specify the DB name.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub < https://github.com/mtxr/vscode-sqltools/issues/110#issuecomment-642410330 , or unsubscribe < https://github.com/notifications/unsubscribe-auth/AAFMX2MFXY5GDCMPZYAI62TRWBRHXANCNFSM4GDBKIQQ

.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/mtxr/vscode-sqltools/issues/110#issuecomment-642628139, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAVP2NBYBWBNW4NPS6N7C4DRWDH7TANCNFSM4GDBKIQQ .

@benyaminl I believe you are using MySQL, right? If so, which version?

Yes I'm using 5.7.27.

Regards, Benyamin Limanto sent from my Sony XPERIA XZ Premium

On Fri, Jun 12, 2020, 06:50 Matheus Teixeira notifications@github.com wrote:

@benyaminl https://github.com/benyaminl I believe you are using MySQL, right? If so, which version?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/mtxr/vscode-sqltools/issues/110#issuecomment-642985928, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAVP2NE7ZUL6CG7C7XQLEF3RWFUS7ANCNFSM4GDBKIQQ .

Hello, I am using your "SQL Tools" extension in visual studio code along with the "Redshift Driver (Dedicated ver)" extension to connect to my data.

When working in a .sql file I first attach the .sql file to my active redshift connection. I am able to run queries within this .sql file, but when typing something as simple as "SELECT * FROM schema_a." no table names are suggested. In other SQL IDE's table names are suggested after typing "schema_a.".

Similarly, after picking the right table, column names are not suggested in the select statement of my query. For example, my query is something like "SELECT tbl_b.* FROM schema_a.table_b tbl_b". Upon typing "tbl_b." no column names are suggested.

I have tried typing ctrl+space to force the table names and column names to show up, but SQL tools returns "Loading..."

Is there a simple solution that I am missing? Thanks in advance!