」工欲善其事,必先利其器。「—孔子《論語.錄靈公》
首頁 > 程式設計 > WordPress REST API無頭表單提交指南

WordPress REST API無頭表單提交指南

發佈於2025-04-15
瀏覽:435

If you’re building a WordPress site, you need a good reason not to choose a WordPress form plugin. They are convenient and offer plenty of customizations that would take a ton of effort to build from scratch. They render the HTML, validate the data, store the submissions, and provide integration with third-party services.

But suppose we plan to use WordPress as a headless CMS. In this case, we will be mainly interacting with the REST API (or GraphQL). The front-end part becomes our responsibility entirely, and we can’t rely anymore on form plugins to do the heavy lifting in that area. Now we’re in the driver’s seat when it comes to the front end.

Forms were a solved problem, but now we have to decide what to do about them. We have a couple of options:

  • Do we use our own custom API if we have such a thing? If not, and we don’t want to create one, we can go with a service. There are many good static form providers, and new ones are popping up constantly.
  • Can we keep using the WordPress plugin we already use and leverage its validation, storage, and integration?

The most popular free form plugin, Contact Form 7, has a submission REST API endpoint, and so does the well-known paid plugin, Gravity Forms, among others.

From a technical standpoint, there’s no real difference between submitting the form‘s data to an endpoint provided by a service or a WordPress plugin. So, we have to decide based on different criteria. Price is an obvious one; after that is the availability of the WordPress installation and its REST API. Submitting to an endpoint presupposes that it is always available publicly. That’s already clear when it comes to services because we pay for them to be available. Some setups might limit WordPress access to only editing and build processes. Another thing to consider is where you want to store the data, particularly in a way that adheres to GPDR regulations.

When it comes to features beyond the submission, WordPress form plugins are hard to match. They have their ecosystem, add-ons capable of generating reports, PDFs, readily available integration with newsletters, and payment services. Few services offer this much in a single package.

Even if we use WordPress in the “traditional” way with the front end based on a WordPress theme, using a form plugin’s REST API might make sense in many cases. For example, if we are developing a theme using a utility-first CSS framework, styling the rendered form with fixed markup structured with a BEM-like class convention leaves a sour taste in any developer’s mouth.

The purpose of this article is to present the two WordPress form plugins submission endpoints and show a way to recreate the typical form-related behaviors we got used to getting out of the box. When submitting a form, in general, we have to deal with two main problems. One is the submission of the data itself, and the other is providing meaningful feedback to the user.

So, let’s start there.

The endpoints

Submitting data is the more straightforward part. Both endpoints expect a POST request, and the dynamic part of the URL is the form ID.

Contact Form 7 REST API is available immediately when the plugin is activated, and it looks like this:

https://your-site.tld/wp-json/contact-form-7/v1/contact-forms//feedback

If we’re working with Gravity Forms, the endpoint takes this shape:

https://your-site.tld/wp-json/gf/v2/forms//submissions

The Gravity Forms REST API is disabled by default. To enable it, we have to go to the plugin’s settings, then to the REST API page, and check the “Enable access to the API” option. There is no need to create an API key, as the form submission endpoint does not require it.

Update (Sep. 10, 2024): The IDs are hashed in Contact Form 7 version 5.8, but can still to be found in the URL of the form’s editing page.

The body of the request

Our example form has five fields with the following rules:

  • a required text field
  • a required email field
  • a required date field that accepts dates before October 4, 1957
  • an optional textarea
  • a required checkbox

For Contact Form 7’s request’s body keys, we have to define them with the form-tags syntax:

{
  "somebodys-name": "Marian Kenney",
  "any-email": "[email protected]",
  "before-space-age": "1922-03-11",
  "optional-message": "",
  "fake-terms": "1"
}

Gravity Forms expects the keys in a different format. We have to use an auto-generated, incremental field ID with the input_ prefix. The ID is visible when you are editing the field.

{
  "input_1": "Marian Kenney",
  "input_2": "[email protected]",
  "input_3": "1922-03-11",
  "input_4": "",
  "input_5_1": "1"
}

Submitting the data

We can save ourselves a lot of work if we use the expected keys for the inputs’ name attributes. Otherwise, we have to map the input names to the keys.

Putting everything together, we get an HTML structure like this for Contact Form 7:

In the case of Gravity Forms, we only need to switch the action and the name attributes:

Since all the required information is available in the HTML, we are ready to send the request. One way to do this is to use the FormData in combination with the fetch:

const formSubmissionHandler = (event) => {
  event.preventDefault();

  const formElement = event.target,
    { action, method } = formElement,
    body = new FormData(formElement);

  fetch(action, {
    method,
    body
  })
    .then((response) => response.json())
    .then((response) => {
      // Determine if the submission is not valid
      if (isFormSubmissionError(response)) {
        // Handle the case when there are validation errors
      }
      // Handle the happy path
    })
    .catch((error) => {
      // Handle the case when there's a problem with the request
    });
};

const formElement = document.querySelector("form");

formElement.addEventListener("submit", formSubmissionHandler);

We can send the submission with little effort, but the user experience is subpar, to say the least. We owe to users as much guidance as possible to submit the form successfully. At the very least, that means we need to:

  • show a global error or success message,
  • add inline field validation error messages and possible directions, and
  • draw attention to parts that require attention with special classes.

Field validation

On top of using built-in HTML form validation, we can use JavaScript for additional client-side validation and/or take advantage of server-side validation.

When it comes to server-side validation, both Contact Form 7 and Gravity Forms offer that out of the box and return the validation error messages as part of the response. This is convenient as we can control the validation rules from the WordPress admin.

For more complex validation rules, like conditional field validation, it might make sense to rely only on the server-side because keeping the front-end JavaScript validation in sync with the plugins setting can become a maintenance issue.

If we solely go with the server-side validation, the task becomes about parsing the response, extracting the relevant data, and DOM manipulation like inserting elements and toggle class-names.

Response messages

The response when there is a validation error for Contact Form 7 look like this:

{
  "into": "#",
  "status": "validation_failed",
  "message": "One or more fields have an error. Please check and try again.",
  "posted_data_hash": "",
  "invalid_fields": [
    {
      "into": "span.wpcf7-form-control-wrap.somebodys-name",
      "message": "The field is required.",
      "idref": null,
      "error_id": "-ve-somebodys-name"
    },
    {
      "into": "span.wpcf7-form-control-wrap.any-email",
      "message": "The field is required.",
      "idref": null,
      "error_id": "-ve-any-email"
    },
    {
      "into": "span.wpcf7-form-control-wrap.before-space-age",
      "message": "The field is required.",
      "idref": null,
      "error_id": "-ve-before-space-age"
    },
    {
      "into": "span.wpcf7-form-control-wrap.fake-terms",
      "message": "You must accept the terms and conditions before sending your message.",
      "idref": null,
      "error_id": "-ve-fake-terms"
    }
  ]
}

On successful submission, the response looks like this:

{
  "into": "#",
  "status": "mail_sent",
  "message": "Thank you for your message. It has been sent.",
  "posted_data_hash": "d52f9f9de995287195409fe6dcde0c50"
}

Compared to this, Gravity Forms’ validation error response is more compact:

{
  "is_valid": false,
  "validation_messages": {
    "1": "This field is required.",
    "2": "This field is required.",
    "3": "This field is required.",
    "5": "This field is required."
  },
  "page_number": 1,
  "source_page_number": 1
}

But the response on a successful submission is bigger:

{
  "is_valid": true,
  "page_number": 0,
  "source_page_number": 1,
  "confirmation_message": "
Thanks for contacting us! We will get in touch with you shortly.
", "confirmation_type": "message" }

While both contain the information we need, they don‘t follow a common convention, and both have their quirks. For example, the confirmation message in Gravity Forms contains HTML, and the validation message keys don’t have the input_ prefix — the prefix that’s required when we send the request. On the other side, validation errors in Contact Form 7 contain information that is relevant only to their front-end implementation. The field keys are not immediately usable; they have to be extracted.

In a situation like this, instead of working with the response we get, it’s better to come up with a desired, ideal format. Once we have that, we can find ways to transform the original response to what we see fit. If we combine the best of the two scenarios and remove the irrelevant parts for our use case, then we end up with something like this:

{
  "isSuccess": false,
  "message": "One or more fields have an error. Please check and try again.",
  "validationError": {
    "somebodys-name": "This field is required.",
    "any-email": "This field is required.",
    "input_3": "This field is required.",
    "input_5": "This field is required."
  }
}

And on successful submission, we would set isSuccess to true and return an empty validation error object:

{
  "isSuccess": true,
  "message": "Thanks for contacting us! We will get in touch with you shortly.",
  "validationError": {}
}

Now it’s a matter of transforming what we got to what we need. The code to normalize the Contact Forms 7 response is this:

const normalizeContactForm7Response = (response) => {
  // The other possible statuses are different kind of errors
  const isSuccess = response.status === 'mail_sent';
  // A message is provided for all statuses
  const message = response.message;
  const validationError = isSuccess
    ? {}
    : // We transform an array of objects into an object
    Object.fromEntries(
      response.invalid_fields.map((error) => {
        // Extracts the part after "cf7-form-control-wrap"
        const key = /cf7[-a-z]*.(.*)/.exec(error.into)[1];

        return [key, error.message];
      })
    );

  return {
    isSuccess,
    message,
    validationError,
  };
};

The code to normalize the Gravity Forms response winds up being this:

const normalizeGravityFormsResponse = (response) => {
  // Provided already as a boolean in the response
  const isSuccess = response.is_valid;
  const message = isSuccess
    ? // Comes wrapped in a HTML and we likely don't need that
      stripHtml(response.confirmation_message)
    : // No general error message, so we set a fallback
      'There was a problem with your submission.';
  const validationError = isSuccess
    ? {}
    : // We replace the keys with the prefixed version;
      // this way the request and response matches
      Object.fromEntries(
        Object.entries(
            response.validation_messages
        ).map(([key, value]) => [`input_${key}`, value])
      );

  return {
    isSuccess,
    message,
    validationError,
  };
};

We are still missing a way to display the validation errors, success messages, and toggling classes. However, we have a neat way of accessing the data we need, and we removed all of the inconsistencies in the responses with a light abstraction. When put together, it’s ready to be dropped into an existing codebase, or we can continue building on top of it.

There are many ways to tackle the remaining part. What makes sense will depend on the project. For situations where we mainly have to react to state changes, a declarative and reactive library can help a lot. Alpine.js was covered here on CSS-Tricks, and it’s a perfect fit for both demonstrations and using it in production sites. Almost without any modification, we can reuse the code from the previous example. We only need to add the proper directives and in the right places.

Wrapping up

Matching the front-end experience that WordPress form plugins provide can be done with relative ease for straightforward, no-fuss forms — and in a way that is reusable from project to project. We can even accomplish it in a way that allows us to switch the plugin without affecting the front end.

Sure, it takes time and effort to make a multi-page form, previews of the uploaded images, or other advanced features that we’d normally get baked right into a plugin, but the more unique the requirements we have to meet, the more it makes sense to use the submission endpoint as we don’t have to work against the given front-end implementation that tries to solve many problems, but never the particular one we want.

Using WordPress as a headless CMS to access the REST API of a form plugin to hit the submissions endpoints will surely become a more widely used practice. It’s something worth exploring and to keep in mind. In the future, I would not be surprised to see WordPress form plugins designed primarily to work in a headless context like this. I can imagine a plugin where front-end rendering is an add-on feature that’s not an integral part of its core. What consequences that would have, and if it could have commercial success, remains to be explored but is a fascinating space to watch evolve.

最新教學 更多>
  • 版本5.6.5之前,使用current_timestamp與時間戳列的current_timestamp與時間戳列有什麼限制?
    版本5.6.5之前,使用current_timestamp與時間戳列的current_timestamp與時間戳列有什麼限制?
    在時間戳列上使用current_timestamp或MySQL版本中的current_timestamp或在5.6.5 此限制源於遺留實現的關注,這些限制需要對當前的_timestamp功能進行特定的實現。 創建表`foo`( `Productid` int(10)unsigned not ...
    程式設計 發佈於2025-07-12
  • 如何同步迭代並從PHP中的兩個等級陣列打印值?
    如何同步迭代並從PHP中的兩個等級陣列打印值?
    同步的迭代和打印值來自相同大小的兩個數組使用兩個數組相等大小的selectbox時,一個包含country代碼的數組,另一個包含鄉村代碼,另一個包含其相應名稱的數組,可能會因不當提供了exply for for for the uncore for the forsion for for ytry...
    程式設計 發佈於2025-07-12
  • Python環境變量的訪問與管理方法
    Python環境變量的訪問與管理方法
    Accessing Environment Variables in PythonTo access environment variables in Python, utilize the os.environ object, which represents a mapping of envir...
    程式設計 發佈於2025-07-12
  • 如何避免Go語言切片時的內存洩漏?
    如何避免Go語言切片時的內存洩漏?
    ,a [j:] ...雖然通常有效,但如果使用指針,可能會導致內存洩漏。這是因為原始的備份陣列保持完整,這意味著新切片外部指針引用的任何對象仍然可能佔據內存。 copy(a [i:] 對於k,n:= len(a)-j i,len(a); k
    程式設計 發佈於2025-07-12
  • Java中Lambda表達式為何需要“final”或“有效final”變量?
    Java中Lambda表達式為何需要“final”或“有效final”變量?
    Lambda Expressions Require "Final" or "Effectively Final" VariablesThe error message "Variable used in lambda expression shou...
    程式設計 發佈於2025-07-12
  • 如何有效地選擇熊貓數據框中的列?
    如何有效地選擇熊貓數據框中的列?
    在處理數據操作任務時,在Pandas DataFrames 中選擇列時,選擇特定列的必要條件是必要的。在Pandas中,選擇列的各種選項。 選項1:使用列名 如果已知列索引,請使用ILOC函數選擇它們。請注意,python索引基於零。 df1 = df.iloc [:,0:2]#使用索引0和1 ...
    程式設計 發佈於2025-07-12
  • 為什麼我在Silverlight Linq查詢中獲得“無法找到查詢模式的實現”錯誤?
    為什麼我在Silverlight Linq查詢中獲得“無法找到查詢模式的實現”錯誤?
    查詢模式實現缺失:解決“無法找到”錯誤在銀光應用程序中,嘗試使用LINQ建立錯誤的數據庫連接的嘗試,無法找到以查詢模式的實現。 ”當省略LINQ名稱空間或查詢類型缺少IEnumerable 實現時,通常會發生此錯誤。 解決問題來驗證該類型的質量是至關重要的。在此特定實例中,tblpersoon可能...
    程式設計 發佈於2025-07-12
  • 如何使用Python的請求和假用戶代理繞過網站塊?
    如何使用Python的請求和假用戶代理繞過網站塊?
    如何使用Python的請求模擬瀏覽器行為,以及偽造的用戶代理提供了一個用戶 - 代理標頭一個有效方法是提供有效的用戶式header,以提供有效的用戶 - 設置,該標題可以通過browser和Acterner Systems the equestersystermery和操作系統。通過模仿像Chro...
    程式設計 發佈於2025-07-12
  • 對象擬合:IE和Edge中的封面失敗,如何修復?
    對象擬合:IE和Edge中的封面失敗,如何修復?
    To resolve this issue, we employ a clever CSS solution that solves the problem:position: absolute;top: 50%;left: 50%;transform: translate(-50%, -50%)...
    程式設計 發佈於2025-07-12
  • 為什麼不````''{margin:0; }`始終刪除CSS中的最高邊距?
    為什麼不````''{margin:0; }`始終刪除CSS中的最高邊距?
    在CSS 問題:不正確的代碼: 全球範圍將所有餘量重置為零,如提供的代碼所建議的,可能會導致意外的副作用。解決特定的保證金問題是更建議的。 例如,在提供的示例中,將以下代碼添加到CSS中,將解決餘量問題: body H1 { 保證金頂:-40px; } 此方法更精確,避免了由全局保證金重置...
    程式設計 發佈於2025-07-12
  • Go語言垃圾回收如何處理切片內存?
    Go語言垃圾回收如何處理切片內存?
    Garbage Collection in Go Slices: A Detailed AnalysisIn Go, a slice is a dynamic array that references an underlying array.使用切片時,了解垃圾收集行為至關重要,以避免潛在的內存洩...
    程式設計 發佈於2025-07-12
  • 為什麼PYTZ最初顯示出意外的時區偏移?
    為什麼PYTZ最初顯示出意外的時區偏移?
    與pytz 最初從pytz獲得特定的偏移。例如,亞洲/hong_kong最初顯示一個七個小時37分鐘的偏移: 差異源利用本地化將時區分配給日期,使用了適當的時區名稱和偏移量。但是,直接使用DateTime構造器分配時區不允許進行正確的調整。 example pytz.timezone(&#...
    程式設計 發佈於2025-07-12
  • 如何使用組在MySQL中旋轉數據?
    如何使用組在MySQL中旋轉數據?
    在關係數據庫中使用mySQL組使用mySQL組進行查詢結果,在關係數據庫中使用MySQL組,轉移數據的數據是指重新排列的行和列的重排以增強數據可視化。在這裡,我們面對一個共同的挑戰:使用組的組將數據從基於行的基於列的轉換為基於列。 Let's consider the following ...
    程式設計 發佈於2025-07-12
  • 如何在鼠標單擊時編程選擇DIV中的所有文本?
    如何在鼠標單擊時編程選擇DIV中的所有文本?
    在鼠標上選擇div文本單擊帶有文本內容,用戶如何使用單個鼠標單擊單擊div中的整個文本?這允許用戶輕鬆拖放所選的文本或直接複製它。 在單個鼠標上單擊的div元素中選擇文本,您可以使用以下Javascript函數: function selecttext(canduterid){ if(d...
    程式設計 發佈於2025-07-12
  • 如何使用Regex在PHP中有效地提取括號內的文本
    如何使用Regex在PHP中有效地提取括號內的文本
    php:在括號內提取文本在處理括號內的文本時,找到最有效的解決方案是必不可少的。一種方法是利用PHP的字符串操作函數,如下所示: 作為替代 $ text ='忽略除此之外的一切(text)'; preg_match('#((。 &&& [Regex使用模式來搜索特...
    程式設計 發佈於2025-07-12

免責聲明: 提供的所有資源部分來自互聯網,如果有侵犯您的版權或其他權益,請說明詳細緣由並提供版權或權益證明然後發到郵箱:[email protected] 我們會在第一時間內為您處理。

Copyright© 2022 湘ICP备2022001581号-3