How does Xcode's autocompletion algorithm work?
I am always amazed by how fast Xcode picks up the code blocks I want to write. I have looked at some different string matching algorithms but none seems to be working as the one Apple uses in Xcode. I would find it quite interesting what type of algorithm they are using.
Thanks in advance.
The image above shows Xcode "predicting" that UTV should be UITableView
It's not just a simple string searching algorithm. It uses nearest code to the scope, last code you picked with the same shortcut, precompiled codes, codes in frameworks, your own defined codes, and many other stuff sorted by some intelligent definition. Since it's private by apple, we may not know how exactly they achieve this. But this year, they open sourced their LSP repository to bring support for these kind of stuff to other editors, even the VIM in terminal! You can investigate on that if you are interested.
Also there are some projects out there like TabNine witch is the all-language autocompleter. It uses machine learning to provide responsive, reliable, and relevant suggestions trained with over 2million github repositories. You can check that out too if you are interested.
Who knows what exactly programming and tech lead companies are currently using while we are looking for algorithms? Maybe a lot of machine learnings is included and only machines knows the exact algorithms.
I'm new to CP. So far I've played with MiniZinc and I'm finding this a lot of fun.
Before I commit too much, I want to figure out what CP language & solver is best to start with in terms of most active community, actively being developed and used in research, and best learning resources (examples & tutorials).
The options so far seem to be: MiniZinc, ECLiPSe, and or-tools.
Thanks,
MiniZinc is an excelent starting point. If you are interested in more languages I suggest to study Choco and JaCoP, both can be used as Java libraries, and both have very active communities.
Also take a look at OptaPlanner (Java, open source, Apache License):
daily commits
active community contributors that send in Pull Requests
monthly beta releases and 2 final releases per year
350+ page manual and also javadocced
unit tests, integration tests and stress test
enterprise support by Red hat
For the last few months I have been playing with, as you yet mentioned, the ECLiPSe programming language to solve all sorts of problems ranging from planning problems to arithmetic and logic puzzles and I've been nothing but amazed by the simplicity and power of this language. It is based on the Prolog language, so if you have the basic knowledge of Prolog, it will be pretty easy and fast to adapt to the language and its syntax. A bonus (imo) to the language are the integrated logical loops which make it very convenient to access certain data. Also, the website provides very clear documentation and code tutorials for all of its functionalities and a page with implementations to a bunch of well-known existing problems is provided here. These especially were very helpful to me while learning the language.
I'm looking for a way to use OpenCL nicely in Haskell, and found these slides (alternative source) by Benedict Gaster. They mention an impressive “HOpenCL Contextual API” but I can't find anything tangible.
The only thing coming close to the C quasiquotation shown seems to be language-c-quote and its OpenCL-C support ends with the types, it doesn't support the extra keywords.
And accelerate is something completely different, and mainly for CUDA, with the OpenCL backend in early alpha.
Then there's HIPERFIT where no code was posted for a year (but the project is still running), which seems to combine the C quasiquotation and OpenCL, their bindings are even called HOpenCL, but are just a wrapper, nothing to see of the monadic transforms etc.
None of this seems close to finished and ready to build upon…
Any news or other projects I missed?
I was looking for exactly the same thing, and I came across this: https://github.com/bgaster/hopencl
This must be what Benedict Gaster - who is not working for AMD anymore - was talking about. There is not a tremendous amount of activity on the git, but there was an update about 2 months ago, which is still better than a year.
EDIT: Actually J. Garret Morris (the other author of HOpenCL) created a fork: https://github.com/jgbm/hopencl
First seeing you post here now. I'm the author of the HIPERFIT-hopencl package and sort of also a bit responsible for the language-c-quote OpenCL C support. I apologize for the naming confusion and that we now have two hopencl-packages. I have mailed Benedict Gaster and J. Garret Morris about how we resolve that.
What do you find lacking in language-c-quote? Could you give an example of what OpenCL C code that it doesn't handle?
(PS. I'm new here and could not find a way to comment on your post, so I had to post this as an answer - perhaps I just haven't reached the right "clearing level" yet)
Let me setup my question with some info. I'm not in college yet and strictly a hobby programmer. Probably a little more than 2 years ago I got started programming on mac. I started with very simplistic GUI examples with Cocoa and XCode. Long story short, I learned from the top down, first learning objective-c, then venturing into more "low-level" projects where I became better at basic C and even used a few C++ libraries in my existing projects.
What I'm saying is that I've never really done anything outside of an XCode project and occasional iPhone project. I've implemented lots of stuff, algorithms, math, etc. but all within that environment. I look at the world of programming and there is so much out there that's not necessarily a standalone application. It seems to me that the hardest thing is finding out where to start; how to setup the environment. I guess I'm wondering if anyone has any suggestions, projects, tutorials, maybe on setting up environments for different languages on different systems. Web programming, java applets? etc.
On the note of environments, I would be interested in knowing on a more basic level what makes a "development environment." To my basic knowledge, an "environment" combines the language, with the compiler that interprets that language, and contains libraries that provide an API for the language, where the compiled product runs on a certain system. This is my basic concept, but again, I'm here.
Sorry if this question... well... combines too many questions, but any input or guidance is welcome. Thanks in advance for any replies!
Not sure if I understood your question correctly or if this will help you, but here are my (relative newbie) thoughts and rambling:
I've done Java at uni in two different courses, one where we wrote the code in Notepad and then compiled it in command line, in some dubious DOS application, and then two years later when we worked in NetBeans and while NetBeans was a lot better and easier, I learned a lot and was a lot more careful when writing code after the Notepad experience (especially after waiting for several minutes for a compile only to see a message caused by a silly bug).
If you can choose between IDEs, I would read on different blogs, see what people prefer and why and make a choice. The problem is that most of the time, both at uni and at work, you can't choose and have to go with the teachers/managers choose, and make the best of it.
It seems to me that the hardest thing is finding out where to start; how to setup the environment.
I think it would be easiest if you found something that you want to do, and then take small steps and get bits done. I work as a desktop app developer and 3 years ago I set up a wordpress blog for a friend and imported posts and comments from a different blogging platform, with minimal knowledge about everything involved. I started with things that were already done by others and learned how to use them and then slowly tried to fill in the gaps - the comments part wasn't done then, so I had to learn about databases, how I could see them and then write the code that inserted in them, etc.
What I'm trying to say is that if you find something to do (and if you don't have ideas for projects, you can find several posts with ideas here, on SO) and then set goals towards doing that, even if you don't finish it, or your studying takes you in areas you hadn't expected, it will all be useful at some point.
I guess I'm wondering if anyone has any suggestions, projects, tutorials, maybe on setting up environments for different languages on different systems. Web programming, java applets? etc.
This is way too broad a question. If you're doing web programming, you need to set up a web programming environment. At a minimum, you would need an HTTP server. You'd probably also need a relational database. The rest of the web environment would be language dependent.
If you're doing GUI programmng, you would need access to the device or devices (iPhone, Android, etc.) that you want to write programs for.
To my basic knowledge, an "environment" combines the language, with the compiler that interprets that language, and contains libraries that provide an API for the language, where the compiled product runs on a certain system.
That gets you started, yes. You'd want an integrated development environment to write the code. Again, you'd probably need a relational or object oriented database. The rest of the development environment is language dependent.
Closed. This question is opinion-based. It is not currently accepting answers.
Want to improve this question? Update the question so it can be answered with facts and citations by editing this post.
Closed 1 year ago.
Improve this question
Have you ever tried learning a language while on a project? I have, and from my personal experience I can say that it takes courage, effort, time, thinking, lots of caffeine and no sleep. Sometimes this has to be done without choice, other times you choose to do it; if you are working on a personal project for example.
What I normally do in this kind of situation, and I believe everyone does, is "build" on top of my current knowledge of languages, structures, syntax and logic. What I find difficult to cope with, is the difference of integrity in some cases. Some languages offer a good background for future learning and "language study", they pose as a good source of information or a frame of reference and can give a "firm" grasp of what's to come. Other languages form or introduce a new way of thinking and are harder to get used to.
Sometimes you unintentionally think in a specific language and when introduced to a new way of thinking, a new language, can cause confusion or make you get lost between the "borders" of your new and your current knowledge of languages.
What can be a good solution in this case? What should be used to broaden the knowledge of the new language, a new way of thinking, and maintain or incorporate the current knowledge of other languages inside the "borders" of the new language?
I find I need to do a project to properly learn a language, but those can be personal projects. When I learned Python on the job, I first expected (and found) a significant slowdown in my productivity for a while. I read the standard tutorials, coding standards and I lurked on the Python list for a while, which gave me a much better idea of the best practices of the language.
Doing things like coding dojos and stuff when learning a language can help you get a feel for things. I just recently changed jobs and went back to Java, and I spent some time working on toy programs just to get back in the feel for things (I'm also reading Effective Java, 2nd edition as my previous major experience had been with Java 1.4).
I think, in some respects no matter what the impetus for learning the language, you have to start by imitating good patterns in the new language. Whether that means finding a good book, with excellent code examples, good on-line tutorials, or following the lead of a more experienced developer, you have to absorb what it means to write good code in a particular language first. Once you have developed a level of comfort, you can start branching out and and experimenting with alternatives to the patterns that you've learned, looking for ways to apply things you've learned from other languages, but keeping within the "rules" of the language. Eventually, you'll get to the point where you know you can 'break the rules" that you learned earlier because you have enough experience to know when they do/don't apply.
My personal preference, even when forced to learn a new language, is to start with some throw away code. Even starting from good tutorials, you'll undoubtedly write code that later you will look back on and not understand how stupid you could have been. I prefer, if possible, to write as my first foray into a language code that will be thrown away and not come back to haunt me later. The alternative is to spend a lot of time refactoring as you learn more and more. Eventually, you'll end up doing this, too.
I would like to mention ALT.NET here
Self-organizing, ad-hoc community of developers bound by a desire to improve ourselves, challenge assumptions, and help each other pursue excellence in the practice of software development.
So in the spirit of ALT.NET, it is challenging but useful to reach out of your comfort zone to learn new languages. Some things that really helped me are as follows:
Understand the history behind a language or script. Knowing evolution helps a lot.
Pick the right book. Research StackOverflow and Amazon.com to find the right book to help you ease the growing pains.
OOP is fairly common in most of the mature languages, so you can skip many of the chapters related to OOP in many books. Syntax learning will be a gradual process. I commonly bookmark some quick handy guides for that.
Read as many community forums as possible to understand the common pitfalls of the new language.
Attend some local meetups to interact with the community and share your pains.
Take one pitch at a time by building small not so complicated applications and thereby gaining momentum.
Make sure you create a reference frame for what you need to learn. Things like how security, logging, multithreading are handled.
Be Open minded, you can be critical, but if you hate something then do not learn that language.
Finally, I think it is worthwhile to learn one strong languages like C# or Java, one functional language and one scripting language like ruby or python.
These things helped me tremendously and I think will help all software engineers and architects to really gear for any development environment.
I learned PHP after I was hired to be the project lead on the Zend Framework project.
It helped that I had 20 years of professional programming background, and good knowledge of C, Java, Perl, JavaScript, SQL, etc. I've also gravitated towards dynamic scripting languages for most of my career. I've written applications in awk, frameworks in shell, macro packages in troff, I even wrote a forum using only sed.
Things to help learn a language on the job:
Reading code and documentation.
Listening to mailing lists and blogs of the community.
Talking to experts in the language, fortunately several of whom were my immediate teammates.
Writing practice code, and asked for code reviews and coaching.(Zend_Console_Getopt was my first significant PHP contribution).
Learning the tools that go along with the language. PHPUnit, Xdebug, phpDoc, phing, etc.
Of course I did apply what I knew from other programming languages. Many computer science concepts are language-universal. The differences of a given language are often simply idiomatic, a way of stating something that can be done another way in another language. This is especially true for languages like Perl or PHP, which both borrow a lot of idioms from earlier languages.
It also helped that I took courses in Compiler Design in college. Having a good foundation in how languages are constructed makes it easier to pick up new languages. At some level, they're all just ways of abstracting runtime stacks and object references.
If you're a junior member of the team and don't know the language, this is not necessarily an issue at all. As long as there is some code review and supervision, you can be a productive.
Language syntax is one issue, but architectural differences are a more important concern. Many languages are also development platforms, and if you don't have experience with the platform, you don't know how to create a viable solution architecture. So if you're the project lead or working solo, you'd better have some experience on the platform before you do your design work.
For example, I would say an experienced C# coder with no VB experience would probably survive a VB.NET project just fine. In fact, it would be more difficult for a developer who only had experience in C#/ASP.NET to complete a C# WPF project than a VB ASP.NET project. An experienced PHP developer might hesitate a bit on a ColdFusion project, but they probably won't make any serious blunders because they are familiar with a script based web development architecture.
Many concepts, such as object modelling and database query strategies, translate just fine between languages. But there is always a learning curve for a new platform, and sometimes it can be quite nasty. The worst case is that the project must be thrown out because the architure is too wrong to refactor.
I like to learn a new language while working on a project, because a real project will usually force me to learn aspects of the language that I might otherwise skip. One of the first things I like to to is read code in that language, and jump in. I find resources (such as books and various internet sites) to help as I go along.
Then, after I've been working on it for a while, I like to read (or re-read) books or other resources on the language. By this time I have some knowledge, so this will help solidify some things and also point out areas where I am flat-out wrong in my understanding. For instance, I can see that I was making incorrect assumptions about similarities between languages.
This also applies to tools -- after using a tool for a while and learning the basics, reading (or skimming) the documentation can teach me a lot.
In my opinion, you should try to avoid that. I know, most of the times you can't but in any case try not to mix the new language with the old one, and never add to the mixture old habits, practices and patterns.
Always try to find resources that will help you get through the new language in the way the language works, not in the way other languages do; that will never have a happy ending, and if it does it will be very hard to modify it to the right way.
Cheers.
Yes I have.
I mean, is there another way? The only language I ever learned that was not on a project was ABC basic, which was what you used on my first computer.
I would recommend if you start with a certain language, stick with it. I only say that because many times in the past I tried more and more different ones, and the one I started out with was the best :D
Everytime I have/want to learn a new language, I force myself to find something to code.
But to be sure I did it well, I always want to be able to check my code and what it ouputs.
To do so, I just try to do the same kind of stuff with languages I know and to compare the outputs. For that, I created a little project (hosted on Github) with an exercise sheet and the correction for every language I learnt. It's a good way to learn in my opinion because it gives you a real little project.