feat(bme280): Implement _ensure_data() auto-trigger pattern by nedseb · Pull Request #318 · steamicc/micropython-steami-lib

nedseb · 2026-03-29T11:59:19Z

Closes #159

Summary

_is_sleep_mode(): reads ctrl_meas register and checks if mode bits are 0x00 (sleep)
_ensure_data(): if the sensor is in sleep mode, triggers a forced measurement and waits for completion; no-op in normal mode
Auto-trigger in measurement methods: temperature(), pressure_hpa(), humidity(), and read() now call _ensure_data() before reading registers, so they always return fresh data regardless of the current operating mode
read_one_shot() is unchanged — it explicitly triggers a forced measurement regardless of mode
6 new mock tests: _is_sleep_mode() in sleep/normal, _ensure_data() trigger/no-op, temperature() and read() auto-trigger verification

Test plan

45 mock tests pass (make test-bme280)
Hardware validation

Copilot

Pull request overview

Adds an auto-trigger “ensure fresh data” pattern to the BME280 driver so measurement APIs can transparently perform a forced conversion when the device is in sleep mode (aligning with the repo’s broader auto-trigger conventions).

Changes:

Add _is_sleep_mode() and _ensure_data() helpers to detect sleep mode and trigger/wait for a forced measurement.
Update temperature(), pressure_hpa(), humidity(), and read() to call _ensure_data() before reading sensor registers.
Add mock scenario tests in tests/scenarios/bme280.yaml to validate sleep/normal detection and auto-trigger behavior.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 4 comments.

File	Description
lib/bme280/bme280/device.py	Introduces `_ensure_data()`/`_is_sleep_mode()` and wires auto-trigger into measurement methods.
tests/scenarios/bme280.yaml	Adds mock scenarios covering sleep/normal mode checks and auto-triggered measurement writes.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-03-29T12:02:45Z

+    def _ensure_data(self):
+        """Trigger a forced measurement if the sensor is in sleep mode.
+
+        In normal mode this is a no-op. In sleep mode it triggers a
+        single conversion and waits for completion so that subsequent
+        register reads return fresh data.
+        """
+        if self._is_sleep_mode():
+            self.trigger_one_shot()
+            self._wait_measurement()


_ensure_data() relies on _wait_measurement(), which currently raises BME280Error("BME280 measurement timeout"). The auto-trigger convention in issue #159 calls for raising OSError on timeout; as implemented, callers will see a driver-specific exception type instead. Consider changing the timeout exception to OSError (or making BME280Error inherit from OSError) so timeout handling is consistent with the convention across drivers.

The project convention uses driver-specific exceptions (e.g. WSENPADSTimeout in wsen-pads, BME280Error here), all inheriting from Exception, not OSError. This is consistent across all drivers in the repo. Keeping BME280Error for now to stay aligned.

You were right — fixed in 3e6fd4f. The convention from #44 is indeed OSError for timeouts, and 5 out of 7 drivers already follow it. I had incorrectly aligned with WSEN-PADS/WSEN-HIDS which are the two outliers. Created #319 to harmonize those two remaining drivers.

Copilot · 2026-03-29T12:02:45Z

    def read(self):
-        """Return (temperature_c, pressure_hpa, humidity_rh) tuple."""
+        """Return (temperature_c, pressure_hpa, humidity_rh) tuple.
+
+        If the sensor is in sleep mode, a forced measurement is triggered
+        automatically before reading.
+        """
+        self._ensure_data()
        raw_temp, raw_press, raw_hum = self._read_raw()


read_one_shot() calls read(), and read() now calls _ensure_data(). On real BME280 hardware, a forced measurement typically returns the sensor to sleep mode after completion; that would make read() immediately trigger a second forced measurement, so read_one_shot() could perform two conversions instead of one. To avoid double-triggering, consider having read_one_shot() read/compensate directly (bypassing _ensure_data), or add an internal/read parameter that skips auto-trigger when read_one_shot() has already triggered a conversion.

Fixed in d39f467: read_one_shot() now reads and compensates directly instead of calling read(), avoiding the double-trigger. A new test "read_one_shot does not double-trigger" verifies that exactly one forced mode write occurs.

Copilot · 2026-03-29T12:02:45Z

+      result = triggered and abs(t - 25.08) < 0.1
+    expect_true: true
+    mode: [mock]


The new auto-trigger coverage only verifies temperature() and read() behavior in sleep mode. Since pressure_hpa() and humidity() were also updated to call _ensure_data(), consider adding similar mock assertions for those methods (and/or asserting p and h in the read() auto-trigger test) to ensure all updated APIs are covered.

Suggested change

result = triggered and abs(t - 25.08) < 0.1

expect_true: true

mode: [mock]

result = (

triggered

and abs(t - 25.08) < 0.1

and abs(p - 1009.21) < 0.1

and abs(h - 50.57) < 0.1

)

expect_true: true

mode: [mock]

- name: "pressure_hpa() auto-triggers in sleep mode"

action: script

script: |

dev.power_off()

i2c.clear_write_log()

p = dev.pressure_hpa()

log = i2c.get_write_log()

triggered = any(reg == 0xF4 and (data[0] & 0x03) == 0x01 for reg, data in log)

result = triggered and abs(p - 1009.21) < 0.1

expect_true: true

mode: [mock]

- name: "humidity() auto-triggers in sleep mode"

action: script

script: |

dev.power_off()

i2c.clear_write_log()

h = dev.humidity()

log = i2c.get_write_log()

triggered = any(reg == 0xF4 and (data[0] & 0x03) == 0x01 for reg, data in log)

result = triggered and abs(h - 50.57) < 0.1

expect_true: true

mode: [mock]

Fixed in d39f467: added auto-trigger tests for pressure_hpa() and humidity(), and the read() test now also validates pressure and humidity values.

Copilot · 2026-03-29T12:02:46Z

+      i2c.clear_write_log()
+      dev._ensure_data()
+      log = i2c.get_write_log()
+      # In normal mode, only a read of ctrl_meas (no write to 0xF4)


This comment says "In normal mode, only a read of ctrl_meas (no write to 0xF4)", but the mock I2C write log does not record reads—so the log will be empty here. Consider rewording to avoid implying the read is observable in the write log.

Suggested change

# In normal mode, only a read of ctrl_meas (no write to 0xF4)

# In normal mode, _ensure_data must not write to ctrl_meas (0xF4); the write log may be empty

Fixed in d39f467: reworded comment to clarify that the write log should not contain a forced mode write.

…sts.

# [0.7.0](v0.6.0...v0.7.0) (2026-03-29) ### Features * **bme280:** Implement _ensure_data() auto-trigger pattern ([#318](#318)) ([a0d344d](a0d344d))

semantic-release-updater · 2026-03-29T19:11:02Z

🎉 This PR is included in version 0.7.0 🎉

The release is available on:

v0.7.0
GitHub release

Your semantic-release bot 📦🚀

feat(bme280): Implement _ensure_data() auto-trigger pattern.

2416ac3

Copilot AI review requested due to automatic review settings March 29, 2026 11:59

github-actions Bot assigned nedseb Mar 29, 2026

Copilot started reviewing on behalf of nedseb March 29, 2026 11:59 View session

Copilot AI reviewed Mar 29, 2026

View reviewed changes

nedseb added the driver Concerne un driver spécifique label Mar 29, 2026

nedseb added 2 commits March 29, 2026 20:47

fix(bme280): Avoid double-trigger in read_one_shot and add missing te…

d39f467

…sts.

fix(bme280): Use OSError for timeouts per convention established in #44.

3e6fd4f

nedseb mentioned this pull request Mar 29, 2026

fix: Harmonize timeout exceptions to OSError per convention #44. #319

Open

4 tasks

nedseb merged commit a0d344d into main Mar 29, 2026
9 checks passed

semantic-release-updater Bot pushed a commit that referenced this pull request Mar 29, 2026

chore(release): v0.7.0. [skip ci]

70a515c

# [0.7.0](v0.6.0...v0.7.0) (2026-03-29) ### Features * **bme280:** Implement _ensure_data() auto-trigger pattern ([#318](#318)) ([a0d344d](a0d344d))

semantic-release-updater Bot added the released label Mar 29, 2026

nedseb deleted the feat/bme280-ensure-data branch March 29, 2026 19:11

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

feat(bme280): Implement _ensure_data() auto-trigger pattern#318

feat(bme280): Implement _ensure_data() auto-trigger pattern#318
nedseb merged 3 commits intomainfrom
feat/bme280-ensure-data

nedseb commented Mar 29, 2026

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI Mar 29, 2026

Uh oh!

nedseb Mar 29, 2026

Uh oh!

nedseb Mar 29, 2026

Uh oh!

Copilot AI Mar 29, 2026

Uh oh!

nedseb Mar 29, 2026

Uh oh!

Copilot AI Mar 29, 2026

Uh oh!

nedseb Mar 29, 2026

Uh oh!

Copilot AI Mar 29, 2026

Uh oh!

nedseb Mar 29, 2026

Uh oh!

Uh oh!

semantic-release-updater Bot commented Mar 29, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

-      result = triggered and abs(t - 25.08) < 0.1
-    expect_true: true
-    mode: [mock]
+      result = (
+        triggered
+        and abs(t - 25.08) < 0.1
+        and abs(p - 1009.21) < 0.1
+        and abs(h - 50.57) < 0.1
+      )
+    expect_true: true
+    mode: [mock]
+  - name: "pressure_hpa() auto-triggers in sleep mode"
+    action: script
+    script: |
+      dev.power_off()
+      i2c.clear_write_log()
+      p = dev.pressure_hpa()
+      log = i2c.get_write_log()
+      triggered = any(reg == 0xF4 and (data[0] & 0x03) == 0x01 for reg, data in log)
+      result = triggered and abs(p - 1009.21) < 0.1
+    expect_true: true
+    mode: [mock]
+  - name: "humidity() auto-triggers in sleep mode"
+    action: script
+    script: |
+      dev.power_off()
+      i2c.clear_write_log()
+      h = dev.humidity()
+      log = i2c.get_write_log()
+      triggered = any(reg == 0xF4 and (data[0] & 0x03) == 0x01 for reg, data in log)
+      result = triggered and abs(h - 50.57) < 0.1
+    expect_true: true
+    mode: [mock]

	# In normal mode, only a read of ctrl_meas (no write to 0xF4)
	# In normal mode, _ensure_data must not write to ctrl_meas (0xF4); the write log may be empty

Conversation

nedseb commented Mar 29, 2026

Summary

Test plan

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI Mar 29, 2026

Choose a reason for hiding this comment

Uh oh!

nedseb Mar 29, 2026

Choose a reason for hiding this comment

Uh oh!

nedseb Mar 29, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 29, 2026

Choose a reason for hiding this comment

Uh oh!

nedseb Mar 29, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 29, 2026

Choose a reason for hiding this comment

Uh oh!

nedseb Mar 29, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 29, 2026

Choose a reason for hiding this comment

Uh oh!

nedseb Mar 29, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

semantic-release-updater Bot commented Mar 29, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants